The article, “Knowledge Combination: A Cocitation Analysis,” authored by Wenpin Tsai and Chia-hung Wu, and published in The Academy of Management Journal in June 2010, presents a compelling exploration into the fundamental mechanisms of knowledge creation within the scholarly community. The authors articulate a widely held academic perspective that the combination of existing knowledge is a pivotal process through which new insights emerge. This process can manifest as incremental refinements of current knowledge, the discovery of novel applications for established concepts, or the invention of entirely new combinations of existing intellectual assets. Within academia, this integral link between combination and creation is clearly observable in how scholars strategically draw upon a diverse array of prior studies to construct the foundational arguments and evidence for their new publications. This integration of disparate research areas or the reconciliation of seemingly contradictory findings serves to advance the understanding of complex organizational phenomena.
Tsai and Wu introduce cocitation analysis as a robust bibliometric technique to empirically investigate these knowledge combinations. Cocitation occurs when two previously published articles are simultaneously referenced by a subsequent article, signaling that the later work has synthesized or connected knowledge from those two earlier sources. By analyzing these cocitation patterns, researchers can develop a comprehensive map of how knowledge is combined and consequently gain a deeper understanding of a research field’s evolution and progress. The validity of cocitation analysis as a tool for understanding academic research field development has been previously demonstrated through comparative analyses using data from actual scholar contacts and word usage patterns within and among subfields.
The authors specifically apply this methodology to the management field, aiming to address a long-standing debate concerning the degree of consensus among management scholars. They ask whether the field is characterized by a single, interconnected knowledge base or by multiple, largely disconnected bodies of knowledge. This investigation also implicitly touches upon philosophical discussions regarding the desired level of consensus within a scientific paradigm, acknowledging diverse views ranging from valuing high consensus for steady scientific progress to embracing a “laissez-faire” approach that prioritizes diversity.
Methodology in Detail: To conduct their extensive analysis, Tsai and Wu selected The Academy of Management Journal (AMJ) as their primary focus, comparing it with three other prominent management journals: Administrative Science Quarterly (ASQ), Journal of Management (JM), and Organization Science (OS). These journals were chosen due to their broad coverage of empirical work across a wide spectrum of management content areas, including organizational behavior, business policy and strategy, organizational theory, human resources, international management, technology, and entrepreneurship. To reduce analytical “noise,” the study concentrated exclusively on high-impact articles. The authors identified the 100 most influential articles published in each of these four journals between 1981 and 2008, using SSCI citation counts as the metric for influence. The reference data from all articles published in these journals during the same period (1981-2008) were then meticulously examined to identify how these 100 influential articles were cocited. The average citation count for these 400 most influential articles was notably high, at 565.01, reinforcing their significance.
Analyses and Key Findings:
- Overview of Cocitation Pairs: The study identified 4,950 possible cocitation pairs among the 100 target articles in each journal. Between 1981 and 2008, AMJ generated the highest number of actual cocitation pairs (1,366), surpassing ASQ (879), JM (1,068), and OS (1,199). While more publications generally lead to more cocitation pairs, it was noteworthy that OS produced more pairs than JM despite having fewer total publications during the period, suggesting different citation dynamics. The total count of cocitations (the number of times any two of the 100 articles were cocited) was also highest for AMJ (5,202).
- Nonoverlapping Cocitation Pairs: The analysis revealed substantial “nonoverlapping” cocitation pairs between journals, indicating distinctive knowledge combination profiles. For instance, a significant 67% of AMJ‘s cocitation pairs were not found in any ASQ articles. Similarly, 56% of AMJ‘s pairs did not appear in JM articles, and another 56% were not in OS articles. This highlights that each journal, while having some overlap, maintains a somewhat unique way of combining foundational knowledge. A robustness check, standardizing the number of publications across journals, actually showed even larger nonoverlapping areas, suggesting that the consensus among journals might be lower than initially presented.
- Cocitation Networks: To understand the overall structure of knowledge combination beyond mere pairs, Tsai and Wu constructed a cocitation network for each journal, where each node represented one of the 100 influential articles, and a connection indicated a cocitation relationship. AMJ demonstrated the highest cocitation network density (0.27) and highest network centralization (44.84%) among the four journals. This suggests that AMJ‘s network is particularly dense with connections and is dominated by a relatively small number of articles that are frequently cocited with many other works.
- Distinctive Cliques and Learning Opportunities: To pinpoint specific areas of unique knowledge combination, the authors performed a clique analysis on the cocitation networks of AMJ and ASQ. Clique analysis was chosen over trait-based clustering because it identifies densely connected subgroups where every node is connected to every other node within the subgroup. This “fully connected” criterion ensures that the identified cliques represent research areas with intensely linked foundational works.
- Distinctive AMJ Clique: A six-article clique was identified in AMJ that was entirely absent from ASQ‘s cocitation network, meaning no pair from this group had ever been cocited in an ASQ article. This clique included Barney (1991), Dess and Davis (1984), Salancik and Pfeffer (1980), Snow and Hrebiniak (1980), Becker and Gerhart (1996), and Huselid (1995). A closer inspection revealed that most cocitations within this clique were related to strategic human resource management (SHRM). This indicated that AMJ publishes and combines knowledge in SHRM in a distinct manner not mirrored by ASQ.
- Distinctive ASQ Clique: Conversely, a five-article clique was found to be distinctive to ASQ, with no cocitations among these articles appearing in AMJ. This clique comprised Kimberly and Evanisko (1981), Barley (1986), Dougherty (1992), Hackman and Wageman (1995), and Edmondson (1999). This ASQ clique was primarily associated with a topic combining sensemaking and innovation, specifically representing a “routine-disruption view of innovation” where new technology adoption necessitates collective sensemaking and adjustments to organizational structure and behavior. AMJ‘s absence of cocitations in this area suggests it does not combine knowledge in this specific domain as ASQ does.
Discussion and Implications: The findings underscore that while management journals share some common ground, each maintains a “somewhat distinctive knowledge combination profile”. These distinct profiles present valuable “learning opportunities” for journals. By examining how knowledge is uniquely combined in other journals, editors and decision-makers can identify currently neglected research areas, potentially leading to special issue opportunities that broaden their journal’s scope. For individual scholars, cocitation analysis offers insights into publication strategies, helping them to strategically draw from prior research and position their new papers for specific journals. Furthermore, for editors and reviewers, this analysis can serve as a benchmark to assess their journal’s unique research contributions and to identify potential biases in their review processes.
Limitations and Future Research: The authors acknowledge several limitations to their study. First, it focused on only four management journals, and expanding the scope to include more management or discipline-based journals would likely reveal even more intricate patterns of knowledge combination. Second, the analysis was limited to the top 100 most-cited papers. While a robustness check with the top 200 papers showed consistent results, the authors suggest that other significant contributions, such as well-cited books and book chapters, should also be considered foundational works in future cocitation analyses.
Despite these limitations, the study offers an initial framework for understanding how existing knowledge is combined in academic research. It also opens avenues for future research into several fascinating questions: How do distinctive research areas evolve over time? What factors drive their development? Why might certain journals not recognize or accept distinctive areas published elsewhere? What roles do editors and reviewers play in facilitating novel knowledge combination and the emergence of new research areas? Investigating these questions could provide profound insights into the social construction of knowledge within the management field and contribute to the sociology of knowledge.
Reference: Tsai, W., & Wu, C.-H. (2010). Knowledge combination: A cocitation analysis. The Academy of Management Journal, 53(3), 441–450. https://www.jstor.org/stable/25684331