Editorial Free

What is the main purpose of a literature review? This editorial serves as a platform for scholars in the logistics and supply chain management (LSCM) field to reflect on our practices in conducting literature review. The editorial aims to initiate a debate on how literature review can add new knowledge to managers and academic researchers.

Since the introduction of systematic literature review to the management discipline (Tranfield et al., 2003), the LSCM field has published different types of literature review. A literature review should generally “map and assess the relevant intellectual territory in order to specify research question which will further develop the knowledge base” (Tranfield et al., 2003, p. 207). The trouble is there are different views about what counts as “intellectual territory” and “development of the knowledge base”, leading to a proliferation of many types of literature review.

A descriptive literature review is a type of review that has gained popularity. A descriptive literature review typically treats the identification of “topics” studied by existing literature as “mapping of the intellectual territory”. Such a review typically describes “what topics” each published article studied. It relies on a description of “who study what topic” or “what topics have (not) been studied” as an effort to understand the “intellectual territory”. Topics receiving more recent attention are often assumed to represent latest “development of the knowledge base”. Is this an effective way to advance scientific knowledge? Sometimes a descriptive literature review categorizes “topics” being studied by the existing literature into a “framework”. What type of framework such an approach develops? Are these conceptual or theoretical frameworks that can advance the understanding of a phenomenon/problem or advance research? Can this type of descriptive literature review advance knowledge? The answer to these questions is “no”.

The case against descriptive literature review is not new. Some of us must have forgotten the main argument of Tranfield et al. (2003). The lack of critically assessment of evidence is the main reason why Tranfield et al. (2003) introduced the evidence-based approach to replace descriptive literature review. In their words, the use of “singular descriptive accounts” by descriptive literature review leads to a lack of “critical assessment” and a “tolerance to loss of knowledge” (Tranfield et al., 2003, p. 207). Some articles that claim to have applied the Tranfield's method seem to rely on an overly descriptive approach to review the literature instead.

A descriptive literature review typically views “topics being studied” as the unit of analysis or the “intellectual territory”. This is in stark contrast with the needs to assess or understand the knowledge base – what we know about a phenomenon or problem. Moreover, many descriptive literature reviews do not critically assess theoretical assumptions and arguments used by the existing literature. Instead, a descriptive literature review typically describes the following contents:

First, a descriptive literature review reports descriptive statistics about topics, methods, theories, journals, key authors or articles, citations and so on. They tell us the proportions of popular topics being studied, and methods and theories used. They inform us the key journals, articles and authors and the highly cited authors (or articles). A descriptive literature review typically does not assess the relevance or appropriateness of theories or methods used by past studies. The emphasis is rather on mapping the distribution or popularity of topics, theories or methods. The question here is whether highlighting popular topics promotes a popular culture rather than a scientific one? The emphasis on presenting descriptive statistics (i.e. topics) is further encouraged by the advances of topic modeling or similar text mining software. The exciting thing about these tools is that they make it possible to apply natural language processing and even machine learning to map topics. Even though such topic modeling tools can help organize large collections of unstructured texts, they do not have the human intellectual capacity to understand deep reasonings and theoretical arguments. Thus, descriptive statistics and topic modeling alone cannot provide an in-depth understanding of what we know and do not know about a phenomenon.

Second, there is a focus on describing what each author or article did in their studies. A descriptive literature review typically describes “author or article X studied Y topic”, or “author or article X used A theory or B method to study Y topic”. Topics such as procurement, logistics, manufacturing, barriers, drives and so on are often used as “theoretical” categories. Some descriptive literature review also summarizes key findings from the reviewed article. There are two common styles in which such a description is presented. First, the descriptions of what each author/article did or found in separate paragraphs, one paragraph (for each article) after another. This author-centric review has been criticized as failing to synthesize the literature in other fields (Webster and Watson, 2002). This type of review does not relate or integrate similar themes or perspectives because related perspectives are presented in separate paragraphs. Second, some review collates the literature that talks about the same topics together, again presenting one topic (in a paragraph/section) after another. This type of review provides an opportunity to relate or integrate arguments and evidence from different articles belong to a similar theme or perspective. It provides an opportunity to cross-examine conflicting findings, views, assumptions, arguments and perspectives. However, the emphasis here is to describe who or which article studied an aspect within a topic with the aim to conclude which aspects have or have not been studied. This focus on topics rather than theoretical themes or perspectives often does not add new theoretical insights or advance theories. Describing what has been studied within a topic does not help researchers to critically assess the use and usefulness of specific concepts, perspectives, assumptions, data, methods and theoretical arguments against the stock of empirical evidence. Information about what topics have (not) been studied does not provide new insights about the phenomenon being studied.

Third, a descriptive literature review sometimes applies bibliometric analysis to identify highly cited authors/articles and use citation networks to illustrate which authors/articles cited each other. This approach presents an opportunity to question why an article is highly cited and what does it really adds to the knowledge base, but it requires researchers to critically read and understand the chains of reasonings and arguments behind the citations. Not all citations truly reflect new insights are obtained from past studies. The only way to understand who truly contribute to new understanding or theory in the field is to understand the chains of reasonings and arguments that truly count (i.e. add knowledge and not citation count). However, bibliometric analysis is often not being used to understand what new knowledge, argument or evidence are being put forward by cited articles. Bibliometric analysis could have been used as an important first step to assess how meaningful citations facilitate or advance either the theory or understanding of a phenomenon. Instead, there is rather a reliance on the bibliometric software to produce illustrative citation networks or citation counts. The availability of software and tools for searching database, topic modeling and bibliometric analysis serve as important first step for a literature review to understand “intellectual territory” (I meant theoretical understanding here). The question is whether an emphasis on listing highly cited articles or authors promote behaviors such as self or meaningless citation, as opposed to advancing knowledge? However marvelous they are, these tools do not give us reasons to skip the most important step that relies on human brains – a critical assessment of the scientific knowledge and tools.

Fourth, a descriptive literature review often relies on the descriptive statistics and categorization of topics being studied to develop a “framework”. A literature review can be used to identify and understand factors that may play important roles in a phenomenon or problem. There are different views what a “framework” should look like and what its main purpose is. A typical framework developed by descriptive literature review groups factors or topics mentioned by past studies into broad categories such as organizational, technological, social, environment, drivers, barriers and so on. These are not theoretical themes. They do not add or create theoretical insights. Without grounded in a theoretical perspective, these broad topics do not help create a conceptual or theoretical framework. Moreover, such a review seldom explains why and how specific factors or categories influence the phenomenon being studied. There is a focus on “listing” all factors mentioned by the existing literature and grouping them into “convenience” categories, rather than that categories grounded in theories. Thus, a descriptive literature review often misses the opportunities to theorize and understand the roles played by known factors. Failing so, it is impossible to differentiate and select relevant factors for suggesting parsimonious models or frameworks to advance knowledge.

Fifth, a descriptive literature review relies on the above descriptive statistics and description of topics being studied to propose future research agenda. Because the emphasis is on the descriptions of topics, such a review does not assess what is known or unknown about the phenomenon being studied. It often does not critically assess the appropriateness or weakness of the concepts, data, methods and theories used. The suggested future research agenda is often based on the argument that some topics are under studied or some methods/theories are under or not yet being used. It is worrying to see this increasing use of the “lack of study” or “lack of theory/method being used” argument to justify future research agenda. Instead, the identification of knowledge gap requires efforts to address “the tensions that lead to the conceptual theory development efforts” (Carter, 2011, p. 5), not simply topics that have not been studied. Meaningful future research agenda can only be identified by identifying important problems that are less understood (Schmenner et al., 2009) and a critical assessment of the methods, theories, concepts and data used by past studies to understand the problems.

In summary, the emphasis on a narrative or description of past studies can at most inform us about topics being studied, methods or theories used, highly cited articles, who cited who or what topics are linked by citations. This information does not advance knowledge or understanding about a phenomenon or tools used by researchers such as theories. One problem facing the field is “too much theory but not enough understanding”, as Schmenner et al. (2009, p. 343) emphasize, “…we need to focus on what appears to be important to know, what we understand about it, and what we can do to understand it better”. Instead of using a descriptive literature review, the field benefits more from literature review that understands important problems and literature review that challenges, sharpens or develops scientific tools (i.e. theories, concepts, data and methods), so we can better understand the real-world.

The goal of research is to “understand a real-life phenomenon or managerial problem” (Schmenner et al., 2009, p. 342). A descriptive literature review that emphasizes descriptive statistics and descriptions of what topics have been studied does not advance knowledge. To advance knowledge or the understanding of real-world problem or phenomenon, the purpose of a literature review must be aligned to the quest for new knowledge and to improve scientific tools. A literature review advances knowledge by integrating past knowledge. Instead of an emphasis to describe “author or article X studied Y topic”, or “author or article X used A theory or B method to study Y topic”, here the emphasis is to use past studies to describe and understand an event or a series of unfolding events, state of affair, feeling, emotion, thought, action, behavior, structure and situation that are observed to exist or occur. Since there are different studies that may have described or analyzed the phenomenon under studied, the job here is to stitch together the diverse information into a more integrated set of understanding.

A literature review should interrogate knowledge about a real-world phenomenon. One way to move away from a descriptive literature review is cross-examine studies that address a similar topic (e.g. procurement) within a phenomenon (e.g. firms addressing sustainability problems). Instead of broad topics, it is more meaningful to integrate and compare studies that applied the same theoretical perspectives or concepts. Since we are management scholars, it is important to view the phenomenon from the perspective of the managers involved. The phenomenon becomes important when it creates serious problems the managers fail to understand and address. A literature review facilitates debates about the problems and perspectives used by the existing literature to understand a phenomenon. For example, Pagell and Shevchenko (2014) challenge the sustainable supply chain literature that emphasizes on researching unsustainable supply chain, rather than a truly sustainable one. Their review of literature identifies five emphases by past studies that restrict the ambition to truly drive research that supports the creation a truly sustainable supply chain. This is what we need in the field – a debate whether our own thinking, norms, actions and practices truly address important problems that matter to the society.

The review of the literature can focus on two types of understanding. First, the real-world problem as understood by managers (or study subjects). In the case of sustainable procurement or logistics, the review of the literature can focus on understanding how procurement or logistics managers view, understand and address sustainability issues in procurement tasks. This first type of understanding emphasizes facts or empirical evidence. Qualitative studies that rely on interviews, case studies and ethnographies may provide rich descriptions that can be integrated together to better understand different aspects. There are also archival data at the Internet, practitioner magazines, press release and policy announcement that reflect how practitioners perceived specific issues for researcher to triangulate with data presented by the academic journals. Observations breed opportunities to develop new concepts and theories. This type of review of observed data can provide a lot of opportunities to generate research questions, e.g. why the managers think, act and behave in certain manners.

The second type of understanding refers to the problems as understood by the literature (the collective scientific knowledge). This type of understanding emphasizes theoretical understanding created by using scientific tools. The scientific communities use or develop different tools to understand the real-world, e.g. data, method, theory, concept, assumption, etc. The review of the literature can instigate what we understand about a phenomenon e.g. how procurement or logistics managers address sustainability issues when specific theoretical perspectives are used. This drives a focus to assess the scientific tools used to understanding a phenomenon, leading to questions e.g. why a theoretical perspective is useful and relevant as opposed to others, which factors/perspectives provide a better understanding of the manager's thinking and behavior and so on. This is where the “tensions that lead to the conceptual theory development efforts” are identified and debated (Carter, 2011, p. 5), theoretical gaps are identified, and modification to existing scientific tools and new theoretical perspectives are suggested. This is how scientist knowledge is advanced. The literature review by Pagell and Shevchenko (2014) doses just that – as a way forward, they suggest changes in (research) norms, measurement, methods and research questions.

A meaningful literature review corroborates and compares the two types of understanding. By doing so, we can instigate what we researchers understand about a phenomenon as opposed to what managers understand about the phenomenon. Such differences can tell us a lot. They can tell us how we may more effectively inform managers about new insight we create from our research. They can challenge the theoretical lens or tools we used to understand a phenomenon, leading to new ways to advance theories and understanding. Thus, future literature reviews should examine important idiosyncratic such as theoretical boundary, unit of analysis, sources of data, study context, definition and operationalization of constructs and research methods (Durach et al., 2017).

A literature review should contribute to theoretical development (Webster and Watson, 2002). A descriptive literature review is often being criticized as failing to add novel theoretical insights or understanding. This is because the emphasis on describing “who study what”, reporting findings without cross-examining theoretical arguments against empirical evidence and claims of novel contribution. Theoretical advancement depends on the integration and juxtaposition of existing literature bases and/or theories that create unique, novel insights.

A literature review helps understand a phenomenon as the filed matures. In a matured field, a literature review synthesizes/examines accumulated body of knowledge and proposes models or theories to extend the existing research or knowledge base. There will be plenty of empirical evidence. So, an evidence-based approach becomes relevant to cross-examine evidence against theoretical claims (Tranfield et al., 2003). Even though the LSCM field has published some meta-analysis, most of them emphasize confirmation of the main theoretical arguments of mostly grand theories, rather than critically assessing their boundary conditions.

The LSCM literature has also published review of the literature in less matured areas or emerging issues, e.g. the adoption of use of blockchain, artificial intelligence, Internet of Things, supply chain finance and digitalization. These emerging issues present great opportunities to describe the phenomenon, propose conceptual model and suggest potential theoretical foundations. Somehow there is a worrying trend. Often, the earlier literature focuses on suggesting the potential benefits of these emerging technologies, even though there is no substantial evidence to support such a claim. When such a literature is reviewed, some reviews describe the benefits as if they are real. We often see statements that claim a specific technology can provide a specific benefit or the technology can create a specific performance outcome. Even though there may be some patchy evidence, such technologies are still rather new and not yet widely adopted. So, the evidence, if any, is often anecdotal in nature. Thus, there is a danger a literature review treats a suggested performance outcome of a new technology as real, and subsequent literature reviews all assume that the performance claim is true. Instead, early evidence is valuable for developing initial understanding, generating new research questions and proposing initial conceptual frameworks, instead of confirming the performance outcome of an emerging technology without examining the evidence.

There are few examples literature review is used as a platform for developing new theories in a field that is still rather new. Even though sustainable operations have been studied since the mid-1990s, the field of sustainable supply chain management was beginning to expand in the 2000s. Carter and Rogers (2008) apply conceptual theory-building methods and the notion of conceptual framework of Meredith (1993, p. 7) to develop a framework, defined as “a collection of two or more interrelated propositions which explain an event, provide understanding, or suggest testable hypotheses”. This can be achieved by integrating different works, summarizing what is common, contrasting differences and extending the existing literature (Meredith, 1993), assessing existing definitions of variables and developing of new variables grounded in appropriate theories. The process involves a combination of induction and “logical deduction” (Carter and Rogers, 2008, p. 362). Here, statements from past studies are treated as qualitative data. They become sources of ideas that form a part of the framework. Note that the review of the literature by Carter and Rogers (2008) does not emphasize descriptive statistics or description of “who study what topic”. They go straight into assessing existing ideas and developing suitable definitions for understanding sustainability and sustainable supply chain management. They suggest a framework by integrating related ideas i.e. risk management, transparency, strategy and culture. They also develop several propositions suggesting ways to achieve sustainable supply chains. This article is not descriptive; it advances theories.

A literature review may also focus on assessing the state and relevance of a theory that has a long history in the field. Schmenner (2001, p. 8) argue “we are duty bound to ferret out the generalizations that endure, and that could be expected to persist well into the 21st century”. Schmenner (2001, p. 89) examines the use of the Theory of Swift, Even Flow published in 1998 (Schmenner and Swink, 1998) to explain productivity gain by “bring together facts and theory”. Schmenner (2001) does not rely on a descriptive literature review. Instead, the article examines different qualitative and quantitative evidence from different countries to conclude several expectations that hold in the past, today and in the future about the Theory of Swift, Even Flow. Off course, one may find new evidence to challenge this theory, given the disruptions caused by COVID-19 pandemic we have seen so far.

Unfortunately, the idea that a theory “lives” forever and it never “die” has become common. Instead, we could use a literature review to systematically show and explain which exiting theories fail to understand certain aspects of a phenomenon. Without doing so, the field will observe slow advances in theory and understanding. For instance, very few scholars in our field question whether resource-based view (RBV) is suitable for understanding LSCM phenomena. Scholars from other fields such as strategic management are more open to cross-examining assumptions and theoretical claims against evidence. Since the introduction of the RBV (Barney, 1991), scholars such as Priem and Butler (2001) have criticized the use of “imprecise definitions” that “hinder prescription” and the use of a static view “relegates causality to a black box” (failing to answer the “how” question). They offer improvement by answering the causal “how” questions and considering the temporal component. Several other scholars have subsequently summarized main problems of RBV and moved on to focus on explaining “how” resources are managed rather than a reliance on using the characteristics or resources to explain firm performance (e.g. Sirmon et al., 2010). Only recently, Bromiley and Rau (2016) put forward some solid reasons why some fundamental assumptions of RBV do not reflect operations and supply chains practices, and that performance explanations offered by RBV do not hold. Our field needs more literature reviews that examine fundamental assumptions among commonly used theories.

The International Journal of Physical Distribution and Logistics Management (IJPDLM) has published many literature reviews in the past. The journal will continue publish literature reviews that add new knowledge and/or advance theories in the LSCM field. Like many journals in the LSCM field, IJPDLM does not encourage the use of descriptive literature reviews that do not contribute much to advancing knowledge or theories. This problem also applies to qualitative or quantitative empirical studies that use descriptive literature reviews to justify knowledge gaps or theoretical contributions. This editorial has outlined some approaches in which a literature review can advance knowledge and theories. It is important to note that software and tools for searching database, topic modeling and bibliometric analysis become useful only when the identified studies are examined or integrated carefully to advance theories or knowledge.

A literature review is an important scientific tool for assessing scientific tools and advancing knowledge. I welcome further efforts to assessing and advancing other types of literature review, especially those that can advance our knowledge/understanding about the real-world and theories.