Purpose

Exaptation is an important source of enterprise innovation, and its antecedents have attracted extensive attention. Drawing on the knowledge ecosystem theory, organizational learning theory, knowledge-based view and dynamic capability theory, this study focuses on how ambidextrous learning (AL) (exploratory and exploitative learning) affects exaptation and examines how knowledge scope (KS) and knowledge reconstruction ability (KRA) moderate the relationship between AL and exaptation.

Design/methodology/approach

Hypotheses were tested via hierarchical linear regression (SPSS) on 354 firm-level questionnaires from Chinese enterprises.

Findings

The results revealed that exploratory and exploitative learning improve exaptation, whereas KS and KRA, respectively, enhance the relationship between AL and exaptation. Further analysis demonstrated that the interaction between KS and KRA significantly strengthened the relationship between exploratory learning and exaptation, while the interaction failed to enhance the relationship between exploitative learning and exaptation.

Practical implications

The study suggests that high-tech firms should engage in both exploratory learning external and exploitative learning activities to promote exaptation. Moreover, leveraging KS and KRA can enhance exaptation.

Originality/value

This study contributes to the literature on AL outcomes by empirically exploring the effects of exploratory and exploitative learning on exaptation, thus extending the application field of the organizational learning theory. Additionally, by revealing the contingency effects of KS and KRA, this study enriching the research on knowledge-based view and dynamic capabilities theory. Furthermore, the research findings can provide helpful implications for enterprises to achieve exaptation effectively.

Exaptation, a novel innovation paradigm, enables cost-efficient industrial transformation and occasionally spawns new markets (Dew et al., 2004; Cattani, 2005). As a strategic mechanism, it has gained prominence in strategic management research for its links to innovative product development (Andriani and Kaminska, 2021), new market creation (Omezzine and Bodas Freitas, 2022), breakthrough innovation (Tang et al., 2022a, b), business model innovation (Codini et al., 2022), niche construction (Dew and Sarasvathy, 2016) and crisis management (Ardito et al., 2021). Meanwhile, China's 2023 Patent Survey reveals that 50% of enterprise patents remain dormant (Nie et al., 2024), signaling severe resource misallocation and stifling endogenous innovation. Exaptation addresses this challenge through its reverse-innovation logic: creating markets from existing technologies (Gould and Vrba, 1982; Ren et al., 2019). By repurposing underutilized technologies firms can activate dormant knowledge (Garud et al., 2018) and bridge the gap between R&D accumulation and commercial application (Ma and Li, 2022). Consequently, elucidating how firms implement strategic actions to achieve exaptation has been a central but unexplored research topic.

Drawing upon the knowledge ecosystem (Por and Molloy, 2000; Yang et al., 2022), exaptation is a process of knowledge flow, linking, coupling and innovation inside and outside the organization (Abatecola et al., 2016), including three links of knowledge input, production and output (Beltagui et al., 2020; Li et al., 2022). To activate exaptation, firms must cultivate diverse knowledge repositories while proactively exploring distant opportunities to trigger functional shifts (Cattani, 2005; Abatecola et al., 2016). Ambidextrous learning (AL) (exploratory vs. exploitative learning) serves as the strategic linchpin (March, 1991). According to the theory of organizational learning, exploratory learning drives novelty-seeking through experimentation and risk-taking (Katila and Ahuja, 2002), whereas exploitative learning focuses on efficiency-driven refinement of existing knowledge (Levinthal and March, 1993). While prior research confirms amidextrous learning impact on product innovation (Duan and Gu, 2021), eco-innovation (Huang et al., 2020), and firm innovation (Tho and Duc, 2021; Cheng et al., 2024), its role in enabling exaptation remains theoretically underexplored. Thus, it is necessary to explore its antecedents and to know how to cultivate exaptation, especially from AL perspective.

Furthermore, within knowledge ecosystems, the transformation of knowledge reserves acquired through AL into exaptation is contingent upon contingency factors. The knowledge-based view posits that applying AL outcomes to novel contexts for exaptation fundamentally depends on knowledge characteristics (Argote, 1999; Capaldo et al., 2017; Ren et al., 2019). As a key attribute, knowledge scope (KS)—defined as the breadth of domains a firm masters (D'Este, 2005)—not only facilitates novel idea acquisition and knowledge recombination (Yayavaram and Ahuja, 2008; Shen et al., 2020) but also enables cross-domain linkages for repurposing existing technologies (Garud et al., 2018). Dynamic capabilities theory further explains how firms integrate, reconfigure, and extend resources to create markets (Eisenhardt and Martin, 2000) and sustain innovative advantage in changing environments (Katkalo et al., 2010; Sun et al., 2019; Antonio et al., 2023). Emerging from the intersection of knowledge-based view and dynamic capabilities theory, knowledge reconstruction ability (KRA)—the capacity to repurpose internal/external knowledge in unpredictable contexts (Ye et al., 2016; Li et al., 2022)—enhances both knowledge utilization efficiency (Wang et al., 2020a, b) and cross-network innovation collisions (Zhong et al., 2024). Based on this, some literatures suggests that exaptation requires coupling internal technical resources with external opportunities (Perry-Smith and Mannucci, 2017), where KS-reconstruction synergy benefits inbound and outbound innovation (Chesbrough and Crowther, 2010; Colombelli et al., 2013). Despite these insights, the tripartite interplay between KS, reconstruction ability, and AL remains underexplored, necessitating systematic investigation of their mechanistic roles in ambidextrous-learning-exaptation relationships.

To address these research gaps, this study builds on the knowledge ecosystem framework, organizational learning theory, the knowledge-based view and dynamic capabilities theory, to examine how exploratory and exploitative learning impact exaptation under the condition of different levels of KS and KRA. Accordingly, the contributions of this study are threefold. First, it introduces knowledge ecosystem to interpret the effectiveness of the theoretical model, expanding the application areas of knowledge ecosystems. Second, Based on the organizational learning theory, the linear impacts of AL on exaptation are analyzed, which expands the scope of ambidextrous literature as well as the research of antecedents of exaptation. Third, this study enriches the knowledge-based view and dynamic capabilities theory literature by investigating the moderating role of KS and KRA in AL outcomes. The study further elucidates the important value of knowledge characteristics in generating exaptation. Overall, this study offers novel insights into the mechanisms of AL into exaptation and the boundary conditions of these mechanisms.

A knowledge ecosystem is defined as a system of tools and methods that connect knowledge producers and users (Kati et al., 2018), focusing on learning, design, and improvement (Por and Molloy, 2000). It addresses critical issues including knowledge production, identification, value creation, renewal, and integration (Borgh et al., 2012), with demonstrated applications in healthcare (Polese et al., 2018), library sciences (Li et al., 2022), and digital transformation (Marinelli et al., 2024). Research indicates that knowledge ecosystems facilitate optimal utilization of internal and external resources, enabling cross-domain knowledge-sharing, linking, interaction, and application to generate unexpected creative ideas (Huston and Sakkab, 2006; Chesbrough and Chen, 2013). Therefore, the process of achieving exaptation, inherently, falls within the domain of knowledge ecosystems. Drawing on the conceptual framework of knowledge ecosystems and the DICE (Distribution, Interaction, Competition and Evolution) model (Chen et al., 2005), this study posits that realizing exaptation requires constructing a knowledge ecosystem encompassing four core processes: First, knowledge acquisition serves as the foundational stage. Under the open innovation paradigm (Chesbrough, 2004), organizational learning theory emphasizes that firms can integrate internal and external knowledge into their systems—either commercializing internal knowledge through outward channels (via exploitative learning) or acquiring external technologies (via exploratory learning). Thus, the establishment of an AL environment (Filippini et al., 2012) is essential for building knowledge repositories and identifying market opportunities for exaptation. Second, knowledge distribution constitutes a critical precondition, directly influencing exaptation outcomes (Garud et al., 2018). The knowledge-based view suggests that exaptation generation is closely tied to knowledge characteristics, which significantly constrain and shape organizational learning processes and effectiveness (Ren et al., 2019). This study operationalizes knowledge distribution through KS, reflecting interdisciplinary, cross-domain diversity. Third, knowledge integration functions as the dynamic capability that ultimately enhances exaptation performance (Garud et al., 2018). Dynamic capabilities theory offers a way for firms to leverage organizational resources and capabilities to achieve performance benefits (Harrigan et al., 2020). KRA, as a quintessential dynamic capability, facilitates the internalization, assimilation, and recombination of knowledge acquired through AL to drive exaptation. This study operationalizes knowledge integration through KRA, by measuring cross-domain linkage and coupling capabilities. Ultimately, the knowledge innovation emerges as the outcome of exaptation, manifesting through the synergistic effects of knowledge acquisition, distribution, and integration—specifically, the interplay between AL, KS, and KRA. Grounded in knowledge ecosystems, this research establishes a theoretical framework that integrates organizational learning theory, the knowledge-based view, and dynamic capabilities theory to examine how AL, KS, and KRA collectively influence exaptation mechanisms.

Exaptation is defined as the process of applying existing technologies to new fields or the ability to leverage these technologies to meet emerging demands in novel environments (Mokyr, 2000; Dew et al., 2004; Andriani and Carignani, 2014), and is also interpreted as existing components that initially have little or additional functionality, but are rearranged and combined into a new, more complex form, giving them new functionality (Gregory, 2008). While attracting extensive attention from practitioners and academia, research on the antecedents of exaptation has become a hot issue (Gregory, 2008; Ren et al., 2019). Gould and Vrba (1982) argue that serendipity serves as the triggering mechanism for exaptation. Furthermore, Cattani (2005) pointed out that the realization of exaptation needs two stages: accumulating knowledge and using opportunities. Garud et al. (2018) identified three key pathways for inducing exaptation: exaptive pools, exaptive events, and exaptive forums. Some recent research emphasizes the external antecedents of exaptation, such as mergers and acquisitions (Dew et al., 2004; Marquis and Huang, 2010), user innovation (Andriani et al., 2017) and institutional interaction (Garud et al., 2018), whereas others highlight internal antecedents, such as knowledge reserves (Cattani, 2005; Abatecola et al., 2016), patent pools (Wang and Zhao, 2023), organizational structure (Andriani et al., 2017), researchers' analogical abilities (Mastrogiorgio and Gilsing, 2016) and multidisciplinary background (Sedita et al., 2022).

Researchers have tried to untangle the antecedents of exaptation, but some research gaps require further discussion. Specifically, prior research has elucidated both endogenous and exogenous factors underlying exaptation along with its generative mechanisms, including exaptive pools, exaptive events, and exaptive forums (Garud et al., 2018). Scholars have further identified the critical role of accumulated knowledge in forming exaptive pools (Dew et al., 2004; Abatecola et al., 2016). However, the specific mechanisms through which firms accumulate diversified knowledge remain unexplored. Based on the organizational learning theory, as a crucial knowledge accumulation approach, the role of AL in exaptation has been less explored. Furthermore, understanding of how exaptation can be generated is limited by considering the role of organizational learning theory. In this vein, exploring AL-exaptation can add new insights to the antecedents of exaptation.

Organizational learning emphasizes that organizations adapt to environmental changes through knowledge acquisition, sharing, and application (Huber, 1991), playing a critical role in knowledge creation and technological innovation (Argote, 1999; Lumpkin and Lichtenstein, 2005; Sanz-Valle et al., 2011; Lee et al., 2015). According to organizational learning theory, exploratory learning is defined as the behavior of organizing to search, acquire and master new knowledge (March, 1991). Exploratory learning drives new technology acquisition and opportunity identification (March, 1991; Dew et al., 2004), thereby enhancing strategic foresight to build heterogeneous knowledge reserves (Cattani, 2005). This process fosters an exaptive pool (Garud et al., 2018), sparks unanticipated technology uses (Chesbrough and Chen, 2013) and creative breakthroughs (Garud et al., 2016), and may support institutional reforms enabling exaptation. In contrast, exploitative learning is defined as the behavior of optimizing organizational processes, mining existing resources and refining existing experience (March, 1991). Exploitative learning enables exaptation through functional repurposing of existing technologies (Gould and Vrba, 1982; Perry-Smith and Mannucci, 2017; Wang et al., 2024a,b), facilitating commercialization in new markets. Thus, we argue that exploratory and exploitative learning positively impact exaptation for the following reasons.

As a key mode of organizational learning, exploratory learning facilitates exaptation by seeking novel methods, technologies, and markets in unfamiliar or unexplored domains (March, 1991; Lin et al., 2015), thereby providing diversified knowledge reserves, market opportunities and change support.

First, enterprises' exploratory learning by means of “industry-university-research cooperation” and “user involvement in innovation” (Chesbrough, 2004; Tang et al., 2022a, b) supplies the knowledge base and potential serendipitous discoveries necessary for exaptation by acquiring new technologies (Cattani, 2005; Andriani et al., 2017). Exploratory learning—through the search, acquisition, and utilization of external heterogeneous technological knowledge—expands inventors' cognitive resources and break through the mindset of “not invented here”, furnishing heterogeneous knowledge pool for exaptation (Chesbrough and Chen, 2013; Li et al., 2024) while also establishing a diversified knowledge foundation for novel combinations of old and new knowledge to generate extra-use exaptation (Weible, 2013; Liu et al., 2025). Moreover, the accumulation of diverse knowledge through exploratory learning enhances firms' sensitivity to market dynamics and industry trends (Feng and Chen, 2022), enabling entrepreneurs to remain alert to contingent events and cultivate unexpected discoveries (Andriani and Carignani, 2014).

Second, exploratory learning triggers exaptation opportunities and fosters innovations that significantly impact the industry (Dew et al., 2004). Lumpkin and Lichtenstein (2005) argue that firms prioritizing exploratory learning are more likely to identify opportunities, while opportunity search and discovery constitute vital phases of exaptation (Cattani, 2005). The existing research shows that exaptation relies more heavily on contexts that diverge from the original functions of existing technologies (Andriani et al., 2017), as more distant knowledge tends to enhance the value of innovation (Capaldo et al., 2017). This suggests that externally acquired heterogeneous knowledge and remote market opportunities obtained through exploratory learning facilitate the emergence of exaptation (Ren et al., 2019). As noted by Cattani and Malerba (2021), exploratory learning enables firms to identify future market trends and seek distant opportunities, thereby assessing potential unexpected functionalities of existing products and technologies in new environments or advancing novel applications of established technologies (Cattani, 2005). Similarly, Haunschild and Sullivan (2002) found that firms engaged in exploratory learning exhibit strategic foresight and anticipatory capabilities, allowing them to access undeveloped yet high-potential markets and domains, thereby securing competitive advantages in new industries based on their existing knowledge base.

Third, exploratory learning challenges enterprises' assumptions about mission, customers, and opportunities, breaks path dependence, and liberates them from institutionalization. Moreover, it facilitates revolutionary cross-border innovation—stimulating internal-external knowledge interaction (Ganzaroli et al., 2014), promoting knowledge reorganization and innovation, and enhancing idea originality (Burgelman and Grove, 2007) and innovative problem-solving (Katila and Ahuja, 2002)—thereby improving the transfer of existing technologies to new fields. Therefore, the following hypothesis was proposed:

H1.

Exploratory learning positively affects exaptation.

As an alternative mode of organizational learning (March, 1991), exploitative learning builds upon existing technological trajectories to enhance internal resource efficiency through refinement, extension, and improvement of established knowledge, thereby maximizing its innovative potential (Levinthal and March, 1993; McCarthy et al., 2017). This study posits that exploitative learning offers an alternative mechanism for enabling exaptation.

First, exploitative learning establishes the technical foundation for exaptation through knowledge outflows (Garud et al., 2016). By continuously mining the potential of existing knowledge and technologies through the path of “technology development”, it facilitates knowledge integration, flow, and dissemination (Ding et al., 2020). On one hand, exploitative learning deepens R&D personnel's the breadth and depth of understanding of product functionalities, break through the constraints of old conventions, enhance creativity, enabling module redesign to create more complex technical forms with novel applications (Gregory, 2008), thus creating technical conditions for functional transfer (Kim and Tuahene-Gima, 2010). On the other hand, through decomposition and refinement of existing components, it expands a technology's exaptive possibilities (Dew et al., 2004; Andriani and Carignani, 2014).

Second, exploitative learning facilitates novel contextual applications of existing technologies (Berchicci, 2013; Andriani et al., 2024). Exploitative learning focuses on continuous in-depth understanding, absorption and expansion of existing products in the form of “technology sharing”, and technologies through refining, integration, extension and transformation, which not only stimulates new perspectives and ways to solve problems, but also helps to promote cross-border thinking and creativity (Caner and Tyler, 2015); But also insight into the new market space with low cost and high efficiency in the way of “ecological co-construction”, and explore the application value of existing technologies in new fields or new scenes through extension (Ghemawat et al., 1993; Jin et al., 2025), thus forming exaptation (Arthur, 2007; Lin et al., 2015). Therefore, the following hypothesis is proposed:

H2.

Exploitative learning positively affects exaptation.

The knowledge-based view emphasizes that knowledge diversity is an important condition for technological innovation (Soto-Acosta et al., 2018; Zhang et al., 2024). The KS refers to the range of domains encompassed by an enterprise's knowledge, playing a crucial role in integrating knowledge across various fields, particularly within complex technological domains (Luo, 2009). This KS facilitates the process of exaptation in enterprises through two primary types of learning activities. On the one hand, a broader KS increases opportunities for association, combination, linkage, and innovation (Katila and Ahuja, 2002; Shoaib and Kehinde, 2019). Specifically, a wide-ranging knowledge base enhances the ability to identify and understand the connections between new and existing knowledge, thus fostering conditions for the cross-disciplinary integration of internal and external insights. Furthermore, a more expansive KS is more likely to yield “happy surprises”, which refers to the unexpected application of concepts from one field to another (Luo, 2009). Conversely, knowledge derived from disparate fields or that deviates from the core business can serve as a significant catalyst for stimulating exaptation (Perry-Smith and Mannucci, 2017). This interplay triggers conditions for exploratory learning-driven exaptation. A notable example is Dr Spencer, an employee at Raytheon Company, who serendipitously discovered that a magnetron could melt chocolate. By integrating knowledge from multiple disciplines—including trigonometry, calculus, chemistry, physics, and metallurgy—he conceptualized the microwave oven and submitted a patent application for the “Method of Cooking Food.” Recognizing the transformative potential of this innovation, Raytheon launched the world's first microwave oven in 1947, revolutionizing cooking methods (Osepchuk, 1984).

On the other hand, a broader scope of knowledge facilitates the effective utilization of existing knowledge reserves. Diverse knowledge spanning various industries, fields, and networks can spark the reconfiguration of products and technologies. This process connects established technological systems with emerging fields and industries, creating opportunities for novel applications of existing products (Chesbrough and Chen, 2013). Such interactions strengthen the relationship between exploitative learning and exaptation. For instance, at Procter and Gamble, a comprehensive knowledge base has expedited the integration of internal discoveries with external ideas, intensifying the development of exaptation (Huston and Sakkab, 2006).

In contrast, a narrow scope of knowledge can constrain an organization's strategy and business practices within traditional or stable domains, leading to a rigidity of core capabilities. This limitation often results in the organization overlooking valuable new knowledge (Beinhocker, 1999). Companies frequently tackle innovation challenges through localized searches, which diminishes their ability to identify opportunities in novel fields via exploratory learning (Feng and Chen, 2022), thereby hindering the likelihood of exaptation. Furthermore, under conditions of limited KS, strategic foresight is often inadequate, and there is a lack of heightened awareness regarding unexpected events. This deficiency makes it challenging to anticipate potentially transformative new environments based on accumulated knowledge (Garud et al., 2016), which is detrimental to the formation of exaptation. Additionally, the narrowness of knowledge domains restricts the feasibility of exploitative learning, impeding the interaction and integration of knowledge across different areas to foster exaptation. Therefore, the following hypotheses were proposed:

H3.

Knowledge scope strengthens the relationship between exploratory learning and exaptation.

H4.

Knowledge scope strengthens the relationship between exploitative learning and exaptation.

KRA refers to an enterprise's capacity to integrate, configure, and innovate internal/external knowledge in response to environmental changes (Wang et al., 2020a, b). Strong capability enables firms to adapt to market trends, overcome inertia, and embed new knowledge into existing systems, forming interconnected knowledge elements (Ye et al., 2016) that drive learning activities toward exaptation. From the perspective of dynamic capabilities theory, knowledge reconstruction bridges internal and external knowledge. Enhanced capability allows firms to filter valuable resources efficiently, amplifying learning's impact on exaptation. Specifically, agile organizations can rapidly identify and leverage critical knowledge, while integrating heterogeneous technologies from exploratory learning into current systems (Bruno et al., 2018). This fosters knowledge recombination and functional transfer, enabling firms to transcend spatial and cognitive boundaries (Andriani and Carignani, 2014), thereby discovering distant opportunities. For example, Huaxi Biotechnology restructured intelligent design, chemistry, and pressure-cooker knowledge using water machine principles, launching a high-concentration hydrogen-rich water machine. This innovation addressed modern families' health and convenience needs while achieving technological breakthroughs.

KRA enhances exaptation through internal knowledge mining and creative integration. First, it strengthens R&D teams' analogical reasoning and technological comprehension (Mastrogiorgio and Gilsing, 2016), enabling them to decompose and integrate technologies while uncovering hidden functional interdependencies (Andriani and Carignani, 2014). Second, it improves internal knowledge search efficiency and predicts the viability of knowledge combinations, facilitating the fusion of new and existing knowledge (Yu and Yu, 2024). For example, BYD (Build Your Dreams) merged battery R&D expertise with material science to optimize power battery structures, ultimately developing the high-performance “blade battery” with extended durability.

Conversely, weak knowledge reconstruction limits opportunity identification and cross-domain knowledge integration, stifling exaptation. Exaptation thrives at intersections of diverse knowledge (Garud et al., 2016), but low capability restricts bridging internal knowledge with external fields. High path dependence and diminished creativity further impede the processing, decomposition, and transformation of existing technologies, weakening the link between exploitative learning and exaptation. Nokia exemplifies this: its overreliance on Symbian systems and failure to integrate emerging technologies prevented radical innovation amid market shifts, leading to its decline. Therefore, the following hypotheses were proposed:

H5.

Knowledge reconstruction ability strengthens the relationship between exploratory learning and exaptation.

H6.

Knowledge reconstruction ability strengthens the relationship between exploitative learning and exaptation.

Beyond their individual roles, KS (diverse knowledge stocks and differentiated structures) and KRA synergistically shape how AL drives exaptation (Colombelli et al., 2013). A broad KS captures external knowledge spillovers while highlighting combinatorial potential (Feng and Chen, 2022), enabling firms to systematically identify, allocate, and integrate cross-domain resources. This process simultaneously strengthens KRA, which reciprocally enhances firms' capacity to bridge old/new knowledge and generate innovations (Wang et al., 2020a, b). Their interplay forms a self-reinforcing cycle: diverse knowledge stocks fuel reconstruction capabilities, while reconstruction efforts expand KS through novel combinations (Savino et al., 2017).

A wide knowledge spectrum provides structural flexibility for multi-path integration (Wang et al., 2017). When coupled with strong reconstruction ability, organizations can decode complex relationships between existing technologies and emerging insights (Wang and Zhao, 2023), enabling them to assimilate externally acquired knowledge into transformative linkages (Xu et al., 2018). This synergy allows firms to deploy exploratory learning beyond conventional boundaries, forging cross-domain connections (Zeng et al., 2015) that fuel creativity and serendipitous discoveries (Mastrogiorgio and Gilsing, 2016). For instance, unexpected linkages between disparate knowledge domains may reveal latent functions in mature technologies (Cattani, 2005), such as repurposing aerospace materials for medical devices. Conversely, limited KS or weak reconstruction capacity disrupts this dynamic. Narrow knowledge stocks restrict combinatorial possibilities, while poor reconstruction skills hinder the translation of insights into actionable connections. Without this synergy, AL becomes fragmented, reducing exaptation potential. Therefore, the following hypotheses were proposed:

H7.

The interaction between knowledge scope and knowledge reconstruction ability strengthens the relationship between exploratory learning and exaptation.

The synergy between KS and KRA drives iterative reuse and innovation of existing technologies (Savino et al., 2017; Zeng et al., 2015). By mining internal knowledge, firms create synergies with cross-domain expertise, enhancing their capacity to repurpose established technologies into novel applications (Carnabuci and Operti, 2013). Specifically, this interaction enables enterprises to reconfigure product components through exploitative learning, unlocking hidden functionalities (Gregory, 2008). For example, deep integration of knowledge modules allows optimization of product performance—such as upgrading a battery's thermal management system to serve both automotive and energy storage sectors—achieving modular exaptation (Andriani and Carignani, 2014).Therefore, the following hypotheses were proposed:

H8.

The interaction between knowledge scope and knowledge reconstruction ability strengthens the relationship between exploitative learning and exaptation.

The research model is shown in Figure 1.

Figure 1
A path diagram shows eight hypotheses “H 1” through “H 8” connecting learning and knowledge to “Exaptation”.The path diagram shows a theoretical research model detailing eight hypotheses, “H 1” through “H 8”, connecting various learning and knowledge constructs to exaptation. On the left, two rounded rectangular boxes are vertically arranged: “Exploratory learning” at the top and “Exploitative learning” at the bottom. Two main diagonal arrows point toward a box on the right labeled “Exaptation”. The arrow from “Exploratory learning” is labeled “H 1” and the arrow from “Exploitative learning” is labeled “H 2”. At the upper center, two boxes labeled “Knowledge scope” and “Knowledge reconstruction ability” are positioned side-by-side. From “Knowledge scope”, two downward arrows point to the paths: “H 3” points to the “H 1” path, and “H 4” points to the “H 2” path. Similarly, from “Knowledge reconstruction ability”, two downward arrows point to the paths: “H 5” points to the “H 1” path, and “H 6” points to the “H 2” path. At the bottom center, a box labeled “Knowledge scope x Knowledge reconstruction ability” represents an interaction. From this box, two upward arrows indicate further paths: “H 7” points to the “H 1” path and “H 8” points to the “H 2” path.

Research model

Figure 1
A path diagram shows eight hypotheses “H 1” through “H 8” connecting learning and knowledge to “Exaptation”.The path diagram shows a theoretical research model detailing eight hypotheses, “H 1” through “H 8”, connecting various learning and knowledge constructs to exaptation. On the left, two rounded rectangular boxes are vertically arranged: “Exploratory learning” at the top and “Exploitative learning” at the bottom. Two main diagonal arrows point toward a box on the right labeled “Exaptation”. The arrow from “Exploratory learning” is labeled “H 1” and the arrow from “Exploitative learning” is labeled “H 2”. At the upper center, two boxes labeled “Knowledge scope” and “Knowledge reconstruction ability” are positioned side-by-side. From “Knowledge scope”, two downward arrows point to the paths: “H 3” points to the “H 1” path, and “H 4” points to the “H 2” path. Similarly, from “Knowledge reconstruction ability”, two downward arrows point to the paths: “H 5” points to the “H 1” path, and “H 6” points to the “H 2” path. At the bottom center, a box labeled “Knowledge scope x Knowledge reconstruction ability” represents an interaction. From this box, two upward arrows indicate further paths: “H 7” points to the “H 1” path and “H 8” points to the “H 2” path.

Research model

Close modal

This study investigates the impact of AL on exaptation through questionnaire surveys, employing a rigorous three-phase development process. To ensure content validity, we first engaged two associate professors specializing in innovation management and two senior executives with R&D experience to refine ambiguous items. Following this validation, we conducted a pilot test in the Taiyuan Economic and Technological Development Zone, Shanxi Province. We randomly selected five enterprises from the zone's official directory and obtained managerial consent to interview three employees per organization. During these structured interviews, we administered preliminary questionnaires while gathering feedback on comprehension challenges and response accuracy. Subsequent analysis of pilot data and participant insights informed comprehensive revisions to the questionnaire's structure, item wording, and logical flow, ultimately producing a finalized instrument that balances academic rigor with practical accessibility, enabling respondents to provide reliable answers based on their practical understanding of exaptation.

Subsequently, this study employed a large-scale questionnaire survey targeting high-tech enterprises in Shanxi, Jiangsu, Shanghai, and Hebei provinces, focusing on managers, technical personnel, and industry-academia-research collaborators as key respondents. The upper echelons theory (Hambrick and Mason, 1984) posits that managers' cognitive backgrounds influence corporate innovation decisions. As key actors in exaptation, managers hold a comprehensive understanding of technological innovation strategies, learning approaches, and knowledge management mechanisms. From the perspective of the community of practice (Choi et al., 2020), the experiential knowledge of technical personnel plays a critical role in innovation iteration. These individuals are directly involved in technological development and possess extensive practical expertise essential for achieving exaptation. Furthermore, the triple helix model underscores the significant impact of industry-university-government collaboration on innovation (Jiménez et al., 2021). Such collaborators provide resource integration strategies that facilitate knowledge flow and collaborative innovation, thereby enabling exaptation. As China has implemented its innovation-driven development strategy since 2012, traditional industries have undergone significant technological transformation, enabling manufacturing sectors to shift from low-end to mid-high-end production while achieving coordinated progress in digitalization, green technologies, and advanced manufacturing. This transition has led increasing numbers of traditional manufacturers to evolve into high-tech enterprises. However, the critical question of how these enterprises maintain sustainable innovation advantages in uncertain environments has drawn substantial attention from both academia and industry (Wanaswa et al., 2021). This study focuses on high-tech enterprises for two principal reasons. First, compared to traditional manufacturers, high-tech enterprises demonstrate more favorable conditions for open innovation, characterized by stronger AL practices, more fluid knowledge flows, and more standardized management mechanisms (Lv et al., 2023). Second, during this crucial economic transformation period, while these enterprises hold numerous invention patents, their technology transfer efficiency remains suboptimal (Nie et al., 2024), creating substantial pressure to adapt existing technologies for sustainable innovation advantages.

To ensure survey validity, we implemented three key measures: (1) verifying respondents' comprehension of concepts related to product innovation, corporate strategy, and industry development; (2) guaranteeing respondent anonymity to minimize participation concerns; and (3) excluding samples containing logical inconsistencies, incomplete responses, or uniform answers.

This study implemented a dual-channel data collection strategy from January to May 2023, combining on-site surveys and online surveys of Master of Business Administration (MBA)/Executive Master of Business Administration (EMBA) participants. Due to resource constraints, we employed convenience sampling, a non-probability method widely adopted in management and innovation studies (e.g. Zhu et al., 2024). (1) On-site surveys: We leveraged the research team's professional networks, alumni associations, and industry-university-research partnerships. From a compiled enterprise directory, we selected target companies based on accessibility and arranged for research team members or designated managers to distribute questionnaires on-site. This approach resulted in 208 valid responses from 240 distributed questionnaires. (2) Online surveys: The questionnaire was administered to MBA/EMBA students affiliated with the research team's academic networks (including advisors, colleagues, and collaborators) from universities in Shanxi, Hebei, Jiangsu, and Shanghai. These respondents, primarily composed of corporate managers and technical professionals engaged in product innovation or industry-academia collaboration across diverse industries, provided diverse and relevant data for this study. A sample of 190 MBA/EMBA students was selected and received the questionnaire via email, accompanied by detailed completion guidelines. Follow-up reminders were sent to ensure timely responses, yielding 169 valid questionnaires.

From the total 430 distributed questionnaires (240 on-site + 190 electronic), we collected 377 responses. After eliminating 23 incomplete, inconsistent, illegible, or incorrectly completed questionnaires, we retained 354 valid responses (81.78% effective rate). Sample characteristics are presented in Table 1.

Table 1

Sample characteristics (N = 354)

VariableItemsFrequencyPercentage (%)
Firm age (years)5–15 years7822.03
16–30 years22363.00
More than 30 years5314.97
Firm scale (employee)Less than 300246.78
301 and 1,0005314.97
1,001 and 2,0008824.86
2,001 and 3,00013337.57
More than 3,0005615.82
Ownershipstate-owned12936.44
private-owned9526.84
three-funded267.34
collective4512.71
Others5916.67
Industrytextile and apparel287.91
furniture6718.93
food processing9426.55
pharmaceutical manufacturing11833.33
Others4713.28
RegionShanxi11733.05
Jiangsu8223.16
Shanghai5615.82
Hebei9927.97

To ensure sample quality and mitigate potential non-response bias and heterogeneity effects from multi-source sampling, we conducted validity tests. First, using Armstrong and Overton's (1977) extrapolation method, we compared early and late respondents on key dimensions (firm size, age, ownership type) and all study variables. The analysis revealed no significant differences (p > 0.05), suggesting minimal non-response bias (cf. Chi et al., 2022). Second, we assessed potential differences between on-site and online samples. F-tests showed no significant differences (all p > 0.05) in study variables (exploratory learning, exploitative learning, exaptation, KS, and KRA) or control variables (firm age, size, ownership type, and industry distribution). Additional independent samples t-tests between the two groups further confirmed these results (all p > 0.05), justifying the pooling of both samples without threatening the study's validity.

The scales employed in this paper are all derived from well-established scales utilized in existing research. To enhance their applicability, the reference scales underwent meticulous translation and back-translation during the design phase, with question items tailored to the unique characteristics of our study. All constructs were measured using 5-point Likert scales, ranging from 1 (strongly disagree) to 5 (strongly agree), to assess the multi-item constructs under investigation. Table 2 lists the measurement items.

Table 2

Factor analysis and the results of reliability and validity

VariableMeasurement itemFactor load
EL_1 (α = 0.711,
CR = 0.822,
AVE = 0.536)
We collect novel information and ideas that went beyond our current market and technological experience0.767
We effectively disseminates and shares the obtained new knowledge and technologies internally0.739
We collect novel information and ideas that went beyond our current market and technological experience0.722
We can quickly use new knowledge and novel technology to develop new market0.700
EL_2 (α = 0.756,
CR = 0.837,
AVE = 0.508)
We can effectively identify new knowledge in the field of current technology0.763
We can effectively create or externally acquire the required knowledge in the field of current technology0.747
We can effectively share the knowledge created or acquired in the field of current technology0.742
We can effectively integrate and apply the knowledge created or acquired in the field of current technology0.699
Our aim is to search for knowledge and information that we can implement well to ensure productivity0.602
EXAP (α = 0.837,
CR = 0.902,
AVE = 0.754)
A technology of our firm has been successfully applied in another field0.885
A product of our firm has gained new functions in the new environment0.881
We have new functions through the combination of existing product functions0.838
KS (α = 0.717,
CR = 0.845,
AVE = 0.645)
Our firm has the knowledge to enhance product functionality and improve product design0.824
Our firm has product knowledge of various industries and market segments0.824
Our firm combines new ideas from diverse technology or product knowledge0.759
KRA (α = 0.710,
CR = 0.821,
AVE = 0.535)
Our firm can understand market opportunities from multiple perspectives0.750
New products of our firm integrate diversified technologies and functions0.747
Ideas and knowledge of enterprise technological innovation originate from various fields0.712
Our firm has a new understanding of market opportunities0.715

AL: Rooted in March's (1991) seminal work, comprises two dimensions: exploratory learning (EL_1) and exploitative learning (EL_2). Exploratory learning was measured with four items, while exploitative learning included five items. All items were adapted from established scales by Atuahene-Gima and Murray (2007), Chung et al. (2015) and Cao (2017).

Exaptation (EXAP): Andriani et al. (2017) proposed an alternative method for measuring exaptation by analyzing user-driven deviations from original functions using secondary data (e.g. DrugDex, ICD-9-CM). Their quantitative approach provides an effective tool for empirical exaptation research within specific industries. However, measuring exaptation across multiple industries remains an area requiring further exploration (Ren et al., 2019). Recent studies by Tang et al. (2022a, b) and Wang and Zhao (2023) introduced a scale-based method, offering another viable measurement tool for exaptation research. Building on the conceptual foundation of exaptation (Adner and Levinthal, 2002; Arthur, 2007; Gregory, 2008) and following Churchill's (1979) scale development paradigm, this study draws on the works of Tang et al. (2022a, b) and Wang and Zhao (2023) to develop and refine a three-item exaptation scale. Validation results demonstrate strong psychometric properties, with all standardized factor loadings exceeding 0.80 (p < 0.001), composite reliability of 0.902, and average variance extracted of 0.754, confirming the scale's validity.

KS: Based on the research of Feng and Chen (2022), three items were used to measure KS.

KRA: Based on the research of Ye et al. (2016), Wang et al. (2020a, b), four items were used to measure KRA.

Control variables: Aligned with established methodologies (Yi et al., 2018; Feng and Chen, 2022; Zhu et al., 2024), we controlled for firm age (operationalized as establishment years: 5–15, 16–30, >30), size (employee tiers: <300, 301–1,000, 1,001–2,000, 2,001–3,000, >3,000), ownership (state-owned, private, three-funded, collective and others), and industry (textile/apparel, furniture, food processing, pharmaceuticals manufacturing and others).

In terms of validity, discriminant validity was confirmed as the square roots of AVEs (bolded diagonal elements in Table 4) exceeded all corresponding inter-construct correlations. Convergent validity was established with all AVEs surpassing the 0.50 benchmark (range: 0.508–0.754) and factor loadings exceeding 0.6 (Table 2). Confirmatory factor analysis supported the hypothesized five-factor structure, showing superior fit (χ2/df = 2.267; comparative fit index = 0.932; Tucker–Lewis index (TLI) = 0.911; goodness of fit index = 0.923; root mean square error of approximation = 0.060) compared to alternative models (Table 3), further verifying discriminant validity.

Table 3

Validity test

Modelχ2/dfCFITLIGFIRMSEASRMR
One factor6.8610.6330.5870.7510.1290.103
Two factors4.6290.7740.7440.8170.1010.049
Three factors4.3280.7950.7660.8340.0970.047
Four factors2.9440.8890.8630.8960.0740.040
Five factors2.2670.9320.9110.9230.0600.035
Five factors + common method factor1.9880.9540.9300.9420.0530.028

Note(s): Five factors: EL_1, EL_2, EXAP, KS, and KRA; four factors: EL_1+ EL_2, EXAP, KS, and KRA; three factors: EL_1+ EL_2, EXAP, KS + KRA; two factors: EL_1+ EL_2+KS + KRA and EXAP; one factor: EL_1+ EL_2+KS + KRA + EXAP

To assess common method variance, we employed two approaches. First, Harman's single-factor test revealed five components (eigenvalues >1), with the first explaining only 33.098% of variance—below the 50% threshold. Second, following Podsakoff et al. (2003), we compared models with/without a common latent variable. All fit index changes (ΔCFI/TLI <0.03; ΔRMSEA/standardized root mean square residual <0.01) were below recommended thresholds, indicating common method variance is not a significant issue (Table 3).

Descriptive statistics were analyzed using SPSS 20.0, with the results presented in Table 4. Significant correlations were observed among the independent variables (exploratory learning and exploitative learning), moderating variables (KS and KRA), and the dependent variable (exaptation). These findings support the suitability of the data for further regression analysis.

Table 4

Mean, standard deviation, and correlation coefficient

Variable12345
EL_10.732    
EL_20.529**0.713   
EXAP0.276**0.360**0.868  
KS0.532**0.455**0.347**0.803 
KRA0.539**0.432**0.264**0.464**0.731
Mean4.17584.15203.60644.31364.0749
Standard deviation0.55250.55501.05710.55590.5751

Note(s): **Significant at the 0.01 level; *significant at the 0.05 level (two-tailed test); the italic data in the diagonal line is the square root value of each variable AVE

Table 5 presents the regression analysis of the independent and dependent variables. Model 1 examines the impact of control variables on exaptation, while Models 2 and 3 assess the effects of exploratory learning and exploitative learning, respectively. Model 2 tests H1, revealing a significant positive relationship between exploratory learning and exaptation (β = 0.275, p < 0.001), thus supporting H1. Similarly, Model 3 tests H2 and demonstrates that exploitative learning also has a significant positive effect on exaptation (β = 0.361, p < 0.001), confirming H2. These results suggest that both exploratory and exploitative learning contribute to exaptation.

Table 5

Analysis results of the effect of ambidextrous learning on exaptation

VariableExaptation
M1M2M3
Firm age−0.037−0.046−0.051
Firm scale−0.025−0.008−0.021
Ownership−0.067−0.058−0.059
Industry0.020−0.0010.021
EL_1 0.275*** 
EL_2  0.361***
△R20.0070.0820.137
F0.6106.191***11.039***

Note(s): ***Significant at the 0.001 level; **significant at the 0.01 level; *significant at the 0.05 level

Table 6 presents the moderating role of KS in the relationship between AL and exaptation. Models 1–2 test H3, showing that KS positively moderates the exploratory learning-exaptation relationship (β = 0.183, p < 0.01), thereby supporting H3. Similarly, Models 3–4 test H4, demonstrating that KS also positively moderates the exploitative learning-exaptation relationship (β = 0.146, p < 0.01), confirming H4. These results indicate that KS enhances both types of learning's effects on exaptation.

Table 6

Analysis results of the moderating role of the knowledge scope

VariableExaptation
M1M2M3M4
Firm age−0.054−0.064−0.058−0.050
Firm scale0.0150.0100.004−0.006
Ownership−0.073−0.068−0.070−0.069
Industry−0.0020.0020.0100.018
EL_10.113*0.168**  
EL_2  0.253***0.281***
KS0.113***0.333***0.237***0.254***
EL_1 × KS 0.183**  
EL_2 × KS   0.146**
△R21.410.1680.1810.200
F9.473***9.969***12.751***11.285***

Note(s): ***Significant at the 0.001 level; **significant at the 0.01 level; *significant at the 0.05 level

As shown in Table 7, we examine the moderating role of KRA in the AL-exaptation relationship. For H5, Models 1–2 demonstrate that KRA positively moderates the effect of exploratory learning on exaptation (β = 0.159, p < 0.01), supporting H5. Similarly, Models 3–4 testing H6 reveal a significant positive moderating effect of KRA on the exploitative learning-exaptation relationship (β = 0.123, p < 0.05), thereby confirming H6. These findings collectively indicate that knowledge reconstruction capability enhances both exploratory and exploitative learning's contributions to exaptation.

Table 7

Moderating role analysis results of knowledge reconstruction ability

VariableExaptation
M1M2M3M4
Firm age−0.050−0.057−0.054−0.054
Firm scale−0.002−0.009−0.012−0.018
Ownership−0.066−0.072−0.065−0.068
Industry0.0010.0030.0160.017
EL_10.184**0.198**  
EL_2  0.302***0.309***
KRA0.169**0.222***0.136*0.171**
EL_1 × KRA 0.159**  
EL_2 × KRA   0.123*
△R20.1020.1230.1520.166
F6.552***6.928***10.359***9.806***

Note(s): ***Significant at the 0.001 level; **significant at the 0.01 level; *significant at the 0.05 level

The moderating effects relationship of the KS, KRA in AL, and exaptation were plotted according to Toothaker (2017) (see Figures 2–6). As displayed in Figure 2, when the KS was at a low level, exploratory learning was positively correlated with exaptation but not significantly (t = 0.245, p > 0.1). When the KS was at a high level, exploratory learning was positively correlated with exaptation (t = 3.863, p < 0.001), which proved that the positive effect of exploratory learning on exaptation became stronger with the increase in the KS, and the low level of exploratory learning had no positive effect. As displayed in Figure 3, when the KS was at a low level, the slope of the straight line was smaller (t = 2.193, p < 0.05), and when the exploitative learning was at a high level, the slope of the straight line was larger (t = 5.334 p < 0.001). This phenomenon revealed that the positive effect of exploitative learning on exaptation became stronger with the increase in the KS, which had a stronger effect than that with the narrow KS.

Figure 2
A line graph titled “Exaptation” showing how “Low” and “High knowledge scope” interact with “Exploratory learning”.The vertical axis represents “Exaptation” and ranges from 2 to 5 in increments of 0.5 units. The horizontal axis features two primary categories: “Low Exploratory learning” and “High Exploratory learning”. Two lines are shown. The legend at the top identifies the solid line with diamond markers as “Low knowledge scope” and the dashed line with square markers as “High knowledge scope”. The line for “Low knowledge scope” starts at the starting point (Low, 2.48) and moves in a slight downward direction with a steady decreasing point to reach an ending point at (High, 2.45). The line for “High knowledge scope” starts at (Low, 2.8) and moves in a steep upward direction with a significant increasing point to reach an ending point at (High, 3.5). Note: All data numerical values are approximated.

Moderating effect of knowledge scope on the relationship between exploratory learning and exaptation

Figure 2
A line graph titled “Exaptation” showing how “Low” and “High knowledge scope” interact with “Exploratory learning”.The vertical axis represents “Exaptation” and ranges from 2 to 5 in increments of 0.5 units. The horizontal axis features two primary categories: “Low Exploratory learning” and “High Exploratory learning”. Two lines are shown. The legend at the top identifies the solid line with diamond markers as “Low knowledge scope” and the dashed line with square markers as “High knowledge scope”. The line for “Low knowledge scope” starts at the starting point (Low, 2.48) and moves in a slight downward direction with a steady decreasing point to reach an ending point at (High, 2.45). The line for “High knowledge scope” starts at (Low, 2.8) and moves in a steep upward direction with a significant increasing point to reach an ending point at (High, 3.5). Note: All data numerical values are approximated.

Moderating effect of knowledge scope on the relationship between exploratory learning and exaptation

Close modal
Figure 3
A line graph shows the interaction effect between levels of exploitative learning and knowledge scope with “Exaptation”.The vertical axis represents “Exaptation” and ranges from 2 to 5 in increments of 0.5 units. The horizontal axis features two primary categories: “Low exploitative learning” and “High exploitative learning”. Two lines are shown. The legend at the top identifies the solid line with diamond markers as “Low knowledge scope” and the dashed line with square markers as “High knowledge scope”. The line for “Low knowledge scope” starts at the starting point (Low, 2.4) and moves in a slight upward direction with a steady increasing point to reach an ending point at (High, 2.65). The line for “High knowledge scope” starts at the starting point (Low, 2.6) and moves in a steep upward direction with a significant increasing point to reach an ending point at (High, 3.5). Note: All data numerical values are approximated.

Moderating effect of knowledge scope on the relationship between exploitative learning and exaptation

Figure 3
A line graph shows the interaction effect between levels of exploitative learning and knowledge scope with “Exaptation”.The vertical axis represents “Exaptation” and ranges from 2 to 5 in increments of 0.5 units. The horizontal axis features two primary categories: “Low exploitative learning” and “High exploitative learning”. Two lines are shown. The legend at the top identifies the solid line with diamond markers as “Low knowledge scope” and the dashed line with square markers as “High knowledge scope”. The line for “Low knowledge scope” starts at the starting point (Low, 2.4) and moves in a slight upward direction with a steady increasing point to reach an ending point at (High, 2.65). The line for “High knowledge scope” starts at the starting point (Low, 2.6) and moves in a steep upward direction with a significant increasing point to reach an ending point at (High, 3.5). Note: All data numerical values are approximated.

Moderating effect of knowledge scope on the relationship between exploitative learning and exaptation

Close modal
Figure 4
A line graph shows the interaction effect between levels of exploitative learning and knowledge scope on “Exaptation”.The vertical axis represents “Exaptation” and ranges from 2 to 5 in increments of 0.5 units. The horizontal axis features two primary categories: “Low Exploratory learning” and “High Exploratory learning”. Two lines are shown. The legend at the top identifies the solid line with diamond markers as “Low knowledge reconstruction ability” and the dashed line with square markers as “High knowledge reconstruction ability”. The line for “Low knowledge reconstruction ability” starts at the starting point (Low, 2.5) and moves in a slight upward direction with a steady increasing point to reach an ending point at (High, 2.6). The line for “High knowledge reconstruction ability” starts at the starting point (Low, 2.7) and moves in a steep upward direction with a significant increasing point to reach an ending point at (High, 3.4). Note: All data numerical values are approximated.

Moderating effect of reconstruction knowledge ability on the relationship between exploratory learning and exaptation

Figure 4
A line graph shows the interaction effect between levels of exploitative learning and knowledge scope on “Exaptation”.The vertical axis represents “Exaptation” and ranges from 2 to 5 in increments of 0.5 units. The horizontal axis features two primary categories: “Low Exploratory learning” and “High Exploratory learning”. Two lines are shown. The legend at the top identifies the solid line with diamond markers as “Low knowledge reconstruction ability” and the dashed line with square markers as “High knowledge reconstruction ability”. The line for “Low knowledge reconstruction ability” starts at the starting point (Low, 2.5) and moves in a slight upward direction with a steady increasing point to reach an ending point at (High, 2.6). The line for “High knowledge reconstruction ability” starts at the starting point (Low, 2.7) and moves in a steep upward direction with a significant increasing point to reach an ending point at (High, 3.4). Note: All data numerical values are approximated.

Moderating effect of reconstruction knowledge ability on the relationship between exploratory learning and exaptation

Close modal
Figure 5
A line graph shows the effect between levels of exploratory learning and knowledge reconstruction ability.The vertical axis represents “Exaptation” and ranges from 2 to 5 in increments of 0.5 units. The horizontal axis features two primary categories: “Low exploitative learning” and “High exploitative learning”. Two lines are shown. The legend at the top identifies the solid line with diamond markers as “Low knowledge reconstruction ability” and the dashed line with square markers as “High knowledge reconstruction ability”. The line for “Low knowledge reconstruction ability” starts at (Low, 2.45) and moves in a slight upward direction with a steadily increasing point to reach an ending point at (High, 2.8). The line for “High knowledge reconstruction ability” starts at (Low, 2.55) and moves in a steep upward direction with a significant increasing point to reach an ending point at (High, 3.4). Note: All data numerical values are approximated.

Moderating effect of reconstruction knowledge ability on the relationship between exploitative learning and exaptation

Figure 5
A line graph shows the effect between levels of exploratory learning and knowledge reconstruction ability.The vertical axis represents “Exaptation” and ranges from 2 to 5 in increments of 0.5 units. The horizontal axis features two primary categories: “Low exploitative learning” and “High exploitative learning”. Two lines are shown. The legend at the top identifies the solid line with diamond markers as “Low knowledge reconstruction ability” and the dashed line with square markers as “High knowledge reconstruction ability”. The line for “Low knowledge reconstruction ability” starts at (Low, 2.45) and moves in a slight upward direction with a steadily increasing point to reach an ending point at (High, 2.8). The line for “High knowledge reconstruction ability” starts at (Low, 2.55) and moves in a steep upward direction with a significant increasing point to reach an ending point at (High, 3.4). Note: All data numerical values are approximated.

Moderating effect of reconstruction knowledge ability on the relationship between exploitative learning and exaptation

Close modal
Figure 6
A line graph illustrates how the interaction of knowledge scope and reconstruction ability affects “Exaptation”.The vertical axis represents “Exaptation” and ranges from 2 to 5 in increments of 0.5 units. The horizontal axis features two primary categories: “Low Exploratory learning” and “High Exploratory learning”. Two lines are shown. The legend at the top identifies the solid line with diamond markers as “Low knowledge scope times knowledge reconstruction ability” and the dashed line with square markers as “High knowledge scope times knowledge reconstruction ability”. The line for “Low knowledge scope times knowledge reconstruction ability” starts at (Low, 2.9) and moves in a slight downward direction through a steady decreasing point to reach an ending point at (High, 2.58). The line for “High knowledge scope times knowledge reconstruction ability” starts at (Low, 2.6) and moves in an upward direction through a steadily increasing point to reach an ending point at (High, 3.2). Note: All data numerical values are approximated.

Moderating effect of the interaction between knowledge scope and knowledge reconstruction ability on the relationship between exploratory learning and exaptation

Figure 6
A line graph illustrates how the interaction of knowledge scope and reconstruction ability affects “Exaptation”.The vertical axis represents “Exaptation” and ranges from 2 to 5 in increments of 0.5 units. The horizontal axis features two primary categories: “Low Exploratory learning” and “High Exploratory learning”. Two lines are shown. The legend at the top identifies the solid line with diamond markers as “Low knowledge scope times knowledge reconstruction ability” and the dashed line with square markers as “High knowledge scope times knowledge reconstruction ability”. The line for “Low knowledge scope times knowledge reconstruction ability” starts at (Low, 2.9) and moves in a slight downward direction through a steady decreasing point to reach an ending point at (High, 2.58). The line for “High knowledge scope times knowledge reconstruction ability” starts at (Low, 2.6) and moves in an upward direction through a steadily increasing point to reach an ending point at (High, 3.2). Note: All data numerical values are approximated.

Moderating effect of the interaction between knowledge scope and knowledge reconstruction ability on the relationship between exploratory learning and exaptation

Close modal

As displayed in Figure 4, under various KRA levels, exploratory learning and exaptation exhibited a differentiated relationship. When KRA was at a low level, exploratory learning was positive but not significantly correlated with exaptation (t = 1.049, p > 0.1). When KRA was high level, exploratory learning was positively correlated with exaptation (t = 4.196, p < 0.001), which revealed that a high level of KRA could significantly enhance the positive effect of exploratory learning on exaptation, whereas a low level of KRA did not have any positive effect. As displayed in Figure 5, at various KRA levels, exploitative learning and exaptation were both positively related. However, the slope of the straight line was smaller (t = 3.104, p < 0.01) at low KRA levels, and the slope of the straight line was larger (t = 5.782, p < 0.001) at high KRA levels, which indicated that high KRA could enhance the role of exploitative learning in promoting exaptation. Furthermore, this condition was more significant than the moderating role of low-level KRA.

Table 8 further examines the joint moderating effect of KS and KRA on dual learning and exaptation. Model 2 in Table 6 shows that the interaction between KS and KRA has a significant positive moderating effect on exploratory learning and exaptation (β = 0.220, p < 0.05), and hypothesis 7 holds true. The hypothesis that the interaction between KS and KRA has a significant positive moderating effect on exploitative learning and exaptation in Model 4 of Table 8 is not significant (β = −0.052, p > 0.05), and Hypothesis 8 is not supported. Figure 6 shows that when the interaction between KS and KRA is at a low level, exploratory learning is positively correlated with exaptation but not significant (t = 0.096, p > 0.1); when the interaction between KS and KRA is at a high level, exploratory learning is significantly positively correlated with exaptation (t = 2.641, p < 0.01), indicating that the larger the interaction between KS and KRA, the stronger the positive promotion effect of exploratory learning on exaptation, further verifying hypothesis 7.

Table 8

The moderation test of the interaction between knowledge scope and knowledge reconstruction ability on ambidextrous learning and exaptation

VariableExaptation
M1M2M3M4
Firm age−0.067−0.060−0.053−0.057
Firm scale0.0090.007−0.004−0.009
Ownership0.004−0.003−0.073−0.075
Industry−0.078−0.0740.0150.014
EL_10.1120.062  
EL_2  0.252***0.270***
KS0.302***0.277***0.229***0.249***
KRA0.146*0.1240.1060.124*
EL_1 × KS0.1230.121  
EL_1 × KRA0.1070.183*  
KS × KRA 0.081 0.082
EL_2 × KS  0.1160.068
EL_2 × KRA  0.0730.056
EL_1 × KS × KRA 0.220*  
EL_2 × KS × KRA   −0.052
△R20.1610.1700.2100.216
F8.550***7.568***10.141***8.590***

First, we conducted a moderation analysis by adding interaction terms between industry type and AL (i.e. exploratory learning and exploitative learning). As shown in Table 9, neither the exploratory learning × industry interaction nor the exploitative learning × industry interaction was significant (p > 0.10), suggesting that the effects of AL do not vary significantly across industries.

Table 9

The moderating effect of industry on ambidextrous learning-exaptation relationships

Moderating variableIndependent variableDependent variablePLLCIULCI
Industryexploratory learningexaptation0.345−0.2480.087
Industryexploitative learningexaptation0.948−0.1580.169

Note(s): LLCI = Lower Limit of Confidence Interval; ULCI = Upper Limit of Confidence Interval

Second, we performed subsample analyses by dividing the full sample into five industries: textile and apparel, furniture, food processing, pharmaceutical manufacturing, and others. Regression results (Table 10) consistently supported the main hypothesis in all subsamples. Notably, the textile and apparel subsample (n < 30) showed slightly weaker statistical power, with exploratory learning positively correlated with exaptation (β = 0.325, p < 0.1) and exploitative learning having a marginally significant effect (β = 0.432, p < 0.05). These results further validate the cross-industry generalizability of our core conclusions.

Table 10

Robustness test of different industry main assumptions

IndustryMain hypothesisβSig.pSample size
Textile and apparelEL_1→EXAP0.3250.0921.75128
EL_2→EXAP0.4320.0222.44228
FurnitureEL_1→EXAP0.3180.0092.70467
EL_2→EXAP0.3040.0122.57167
Food processingEL_1→EXAP0.2830.0062.83194
EL_2→EXAP0.3810.0003.94894
Pharmaceutical manufacturingEL_1→EXAP0.3690.0004.282118
EL_2→EXAP0.2770.0023.104118
OthersEL_1→EXAP0.3340.0222.38047
EL_2→EXAP0.3290.0242.33447

This study examines the relationship between AL and exaptation, exploring the moderating roles of KS and KRA. Based on data from 354 Chinese firms, our results reveal two key findings. First, we find that both exploratory and exploitative learning have a positive impact on exaptation through dual mechanisms: exploratory learning contributes by acquiring new technologies and markets, thereby establishing both knowledge reservoirs and implementation conditions for novel functionalities of existing technologies in new domains; simultaneously, exploitative learning enhances exaptation likelihood through refining, extending, improving, and transforming established technologies to generate new functions or adapt them to novel environments. These results align with prior research on AL and specific types of innovation. Duan and Gu (2021) suggested that AL can enhance new product innovation, while Huang et al. (2020) explored the effect of AL on eco-innovation, and Cao et al. (2019), Cheng et al. (2024) revealed the relationship between AL and firm innovation. Our study extends these findings by demonstrating the positive effect of both exploratory and exploitative learning on exaptation.

Second, we find that KS moderates the relationship between AL and exaptation, consistent with previous research demonstrating KS's amplification effect on innovation performance through knowledge acquisition (Chung et al., 2018; Li et al., 2021a,b,c; Wang et al., 2024a, b). Specifically, broader knowledge scopes expand domain coverage, thereby enhancing AL's capacity to foster exaptation. Additionally, our study shows that KRA significantly moderates the effects of AL on exaptation, consistent with previous research on the moderating role of KRA in the formation of innovation performance (Wang et al., 2023). This is because stronger reconstruction ability facilitates internal-external knowledge integration and novel function identification. Crucially, the interaction between KS and reconstruction ability positively moderates the relationship between exploratory learning and exaptation by simultaneously providing strategic foresight while establishing knowledge integration foundations. However, this synergistic effect proves insignificant for exploitative learning-exaptation relationships, potentially due to firms' failure to develop integrated mechanisms that combine exploitative learning with KS and reconstruction ability for optimal exaptation outcomes.

This study contributes to the theoretical literature in several ways. First, it advances AL research by establishing its connection with exaptation—an underexplored innovation mechanism critical for technological evolution (Gould and Vrba, 1982; Ren et al., 2019). While existing studies have extensively documented AL's impacts on conventional innovation outcomes, including disruptive innovation (Yang, 2024), product innovation (Duan and Gu, 2021), innovation performance (Cao et al., 2019; Huang et al., 2020), and sustainability metrics (Lv et al., 2021), its role in enabling exaptation remained theoretically underdeveloped. Our findings bridge this gap through empirical validation of AL's exaptation facilitation mechanisms, thereby expanding the theoretical boundaries in organizational learning research.

Second, this study advances the exaptation literature by identifying AL as a novel driver through based on organizational learning theory. While prior research has established exaptation's cross-industry significance (Ren et al., 2019; Liu et al., 2021; Li et al., 2022) and identified key antecedents—including product-environment-user interactions (Andriani et al., 2017), cross-domain knowledge integration (Perry-Smith and Mannucci, 2017), knowledge heterogeneity (Dew et al., 2004; Abatecola et al., 2016), and customer-driven innovation (Andriani et al., 2017; Tang et al., 2022a, b)—existing studies inadequately address systematic knowledge acquisition mechanisms. We bridge this gap by demonstrating how AL synergistically integrates exploratory learning's knowledge-reservoir development for novel applications (McCarthy et al., 2017) with exploitative learning's capability for functional augmentation and environmental adaptation (Levinthal and March, 1993), thereby proposing a more holistic exaptation framework that transcends prior fragmented approaches.

Third, this study advances theoretical understanding by identifying KS and reconstruction ability as critical moderators of the AL-exaptation relationship. Grounded in the knowledge-based view, we demonstrate how broader KS amplifies this linkage, extending prior work on its role in network innovation (Men and Lan, 2023) and brokerage dynamics (Rong et al., 2024). From a dynamic capabilities perspective, we reconceptualize KRA as a significant enhancer (versus its traditional framing as a mediator/predictor in Ye et al., 2016; Marinelli et al., 2024; Luo et al., 2024). Crucially, we uncover a synergistic moderating effect when KS interacts with reconstruction capability, with particularly pronounced effects in exploratory learning contexts. These findings respond to Wang and Zhao's (2023) call for boundary condition investigations while providing a nuanced framework explaining how organizational knowledge architectures shape exaptation outcomes.

Fourth, this study makes significant theoretical contributions by developing a dual-path framework of exaptation generation through synthesizing knowledge ecosystems, organizational learning theory, knowledge-based view, and dynamic capabilities theory. Meanwhile, drawing on Caputo (2024) research, we identify two distinct patterns: (1) The “outside-in” model demonstrates how exploratory learning facilitates exaptation through three sequential stages: acquiring external technological/market knowledge (McCarthy et al., 2017), integrating this knowledge with existing systems through KS and reconstruction capability, and ultimately generating either extra-use exaptation (Weible, 2013) or scenario-driven exaptation (Andriani and Carignani, 2014), as shown in Figure 7. (2) Conversely, the “inside-out” model reveals how exploitative learning transforms existing technologies via refinement and extension, which then adapt to new environments through KS-enabled diversification (Luo, 2009) or capability-driven recombination (Levinthal and March, 1993). By integrating these theoretical perspectives, we provide the first comprehensive framework that systematically explains exaptation generation mechanisms, significantly advancing beyond previous partial examinations of antecedent factors, as shown in Figure 8.

Figure 7
A flowchart shows “Knowledge flow” from “exploratory learning” leading to two types of technological exaptation.The flowchart shows a model of knowledge dynamics and technological exaptation. On the upper left, a rectangular box labeled “exploratory learning” connects downward and then rightward to a central dash dot rectangular container. This path is labeled “Knowledge flow”. Above the central container, two rectangular boxes labeled “knowledge scope” and “knowledge reconstruction ability” point downward with arrows into the central dashed container. The central dashed container is divided into two horizontal rows by a solid line. In the top row, two rounded rectangular boxes labeled “new technologies” and “existing technologies” are joined by a plus symbol. An arrow from this pair points rightward to an oval labeled “extra-use exaptation”. In the bottom row, two rounded rectangular boxes labeled “new Market” and “existing technologies” are joined by a plus symbol. An arrow from this pair points rightward to an oval labeled “scenario driven exaptation”. At the bottom of the diagram, a separate large dash-dot rectangular container holds four rounded rectangular boxes arranged horizontally: “knowledge acquisition”, “knowledge distribution”, “knowledge integration”, and “knowledge innovation”.

Exploratory learning-driven exaptation based on ecosystem theory

Figure 7
A flowchart shows “Knowledge flow” from “exploratory learning” leading to two types of technological exaptation.The flowchart shows a model of knowledge dynamics and technological exaptation. On the upper left, a rectangular box labeled “exploratory learning” connects downward and then rightward to a central dash dot rectangular container. This path is labeled “Knowledge flow”. Above the central container, two rectangular boxes labeled “knowledge scope” and “knowledge reconstruction ability” point downward with arrows into the central dashed container. The central dashed container is divided into two horizontal rows by a solid line. In the top row, two rounded rectangular boxes labeled “new technologies” and “existing technologies” are joined by a plus symbol. An arrow from this pair points rightward to an oval labeled “extra-use exaptation”. In the bottom row, two rounded rectangular boxes labeled “new Market” and “existing technologies” are joined by a plus symbol. An arrow from this pair points rightward to an oval labeled “scenario driven exaptation”. At the bottom of the diagram, a separate large dash-dot rectangular container holds four rounded rectangular boxes arranged horizontally: “knowledge acquisition”, “knowledge distribution”, “knowledge integration”, and “knowledge innovation”.

Exploratory learning-driven exaptation based on ecosystem theory

Close modal
Figure 8
A flowchart “exploitative learning” leading to two types of technological exaptation.The flowchart shows a model of knowledge dynamics and technological exaptation. On the upper left, a rectangular box labeled “exploitative learning” connects downward and then rightward to a central dashed rectangular container. This path is labeled “Knowledge flow”. Above the central container, two rectangular boxes labeled “knowledge scope” and “knowledge reconstruction ability” point downward with arrows into the central dashed container. The central dashed container is divided into two horizontal rows by a solid line. In the top row, two rounded rectangular boxes labeled “existing technologies” and “existing technologies” are joined by a plus symbol. An arrow from this pair points rightward to an oval labeled “extra-use exaptation”. In the bottom row, two rounded rectangular boxes labeled “existing technologies” and “new fields” are joined by a plus symbol. An arrow from this pair points rightward to an oval labeled “Scenario driven exaptation”. At the bottom of the diagram, a separate large dashed rectangular container holds four rounded rectangular boxes arranged horizontally: “knowledge acquisition”, “knowledge distribution”, “knowledge integration”, and “knowledge innovation”.

Exploitative learning-driven exaptation based on ecosystem theory

Figure 8
A flowchart “exploitative learning” leading to two types of technological exaptation.The flowchart shows a model of knowledge dynamics and technological exaptation. On the upper left, a rectangular box labeled “exploitative learning” connects downward and then rightward to a central dashed rectangular container. This path is labeled “Knowledge flow”. Above the central container, two rectangular boxes labeled “knowledge scope” and “knowledge reconstruction ability” point downward with arrows into the central dashed container. The central dashed container is divided into two horizontal rows by a solid line. In the top row, two rounded rectangular boxes labeled “existing technologies” and “existing technologies” are joined by a plus symbol. An arrow from this pair points rightward to an oval labeled “extra-use exaptation”. In the bottom row, two rounded rectangular boxes labeled “existing technologies” and “new fields” are joined by a plus symbol. An arrow from this pair points rightward to an oval labeled “Scenario driven exaptation”. At the bottom of the diagram, a separate large dashed rectangular container holds four rounded rectangular boxes arranged horizontally: “knowledge acquisition”, “knowledge distribution”, “knowledge integration”, and “knowledge innovation”.

Exploitative learning-driven exaptation based on ecosystem theory

Close modal

First, this study provides a dual-path framework for manufacturing exaptation, addressing prior overemphasis on isolated factors like M&A or organizational culture (Dew et al., 2004; Andriani et al., 2017). By integrating AL, KS, and reconstruction ability, firms can sustain serendipitous discoveries and contextualize them into exaptation (Abatecola et al., 2016). This synergy enables two behavioral logics: (1) Inside-out: Leveraging internal knowledge systems to repurpose technologies (e.g. adapting automotive battery tech for energy storage). (2) Outside-in: Absorbing external insights to redefine existing patents (e.g. applying AI-driven data from smart factories to optimize logistics). Both approaches resolve the “function transfer” challenge while building low-cost innovation pathways (Ren et al., 2019).

Second, the framework tackles dormant patent issues plaguing Chinese manufacturers (Yuan, 2009; Nie et al., 2024). Key strategies include: (1) Cross-boundary collaboration: Utilize industry-university partnerships or tech exhibitions to diversify knowledge interactions, activating dormant patents through unexpected combinations (Garud et al., 2018). (2) Ecosystem building: Engage users, communities, and institutions to create inspiration-rich scenarios (Baldwin and von Hippel, 2011). For instance, open innovation platforms stimulate collective inspiration, creativity, and imagination, ultimately facilitating the discovery of new applications for 'sleeping patents' through contextual exploration (Nie et al., 2024). (3) Knowledge restructuring: Bridge external knowledge with internal systems to unlock novel applications (Ye et al., 2016). Procter and Gamble's “Connect + Develop” program exemplifies this, transforming dormant patents into market-ready products via external partnerships (Huston and Sakkab, 2006).

This study has several limitations. First, while it examines the impact of AL (exploratory and exploitative learning) on exaptation, complex internal dynamics—such as how exploratory learning influences exploitative learning, and their complementary or synergistic relationships (Yi et al., 2018)—remain underexplored and warrant future investigation. Second, AL may indirectly drive exaptation through mediating mechanisms like cross-border search or dynamic capabilities. Future studies should test these pathways to clarify causal linkages. Third, the use of cross-sectional data limits insights into temporal dynamics. Since AL effects and exaptation often unfold over time, longitudinal designs are needed to rigorously establish causality (Wang et al., 2012). Fourth, the study of exaptation offers significant potential for expansion in terms of research methodologies and contexts (Ren et al., 2019). Methodologically, prior research on exaptation has predominantly relied on qualitative approaches, with a notable lack of empirical studies. Andriani et al. (2017) proposed an alternative method for measuring exaptation by analyzing user-driven deviations from original functions using secondary data (e.g. DrugDex, ICD-9-CM), providing a theoretical foundation for exaptation measurement in specific industries such as pharmaceuticals. However, cross-industry measurement of exaptation remains an unresolved challenge. Tang et al. (2022a, b) and Wang and Zhao (2023) conducted empirical studies on exaptation using scale-based methods, offering another viable approach for its measurement. Nevertheless, research on exaptation scales is still in its infancy and requires further systematic exploration by scholars to enrich the methodological toolkit (Ren et al., 2019; Wang and Zhao, 2023). In terms of research contexts, exaptation can be examined across diverse industries and cultural settings (Ren et al., 2019). On one hand, prior studies have largely focused on qualitative analyses within specific industries, making cross-industry empirical research a promising direction. On the other hand, most existing research has been conducted in Western developed countries, highlighting the importance of exploring exaptation in the context of Eastern developing nations. Finally, while common method bias was mitigated through questionnaire design (e.g. anonymization and reverse-coding), self-reported data inherently carry subjectivity (Zhu et al., 2024). Complementing surveys with qualitative methods—such as case studies or expert interviews—could strengthen methodological robustness.

Abatecola
,
G.
,
Belussi
,
F.
,
Breslin
,
D.
and
Filatotchev
,
I.
(
2016
), “
Darwinism, organizational evolution and survival: key challenges for future research
”,
Journal of Management and Governance
, Vol. 
20
No. 
1
, pp. 
1
-
17
, doi: .
Andriani
,
P.
and
Carignani
,
G.
(
2014
), “
Modular exaptation: a missing link in the synthesis of artificial form
”,
Research Policy
, Vol. 
43
No. 
9
, pp. 
1608
-
1620
, doi: .
Andriani
,
P.
and
Kaminska
,
R.
(
2021
), “
Exploring the dynamics of novelty production through exaptation: a historical analysis of coal tar-based innovations
”,
Research Policy
, Vol. 
50
No. 
2
, 104171, doi: .
Adner
,
R.
and
Levinthal
,
D.A.
(
2002
), “
The emergence of emerging technologies
”,
California Management Review
, Vol. 
45
No. 
1
, pp.
50
-
66
, doi: .
Andriani
,
P.
,
Ali
,
A.
and
Mastrogiorgio
,
M.
(
2017
), “
Measuring exaptation and its impact on innovation, search, and problem solving
”,
Organization Science
, Vol. 
28
No. 
2
, pp. 
320
-
338
, doi: .
Andriani
,
P.
,
Cattani
,
G.
,
Dang
,
J.R.
and
Kaminska
,
R.
(
2024
), “
From perfume to spirits: exaptation of a bundle of local resources by an outsider entrepreneur
”,
Research Policy
, Vol. 
54
No. 
1
, 105137, doi: .
Antonio
,
N.
,
Pablo
,
L.
,
Eloy
,
G.
and
Maria
,
D.D.
(
2023
), “
Intangible resources, static and dynamic capabilities and perceived competitive advantage in exporting firms, A PLS-SEM/fsQCA approach
”,
Technological Forecasting and Social Change
, Vol. 
198
, 123001.
Ardito
,
L.
,
Coccia
,
M.
and
Petruzzelli
,
A.M.
(
2021
), “
Technological exaptation and crisis management: evidence from COVID-19 outbreaks
”,
R&D Management
, Vol. 
51
No. 
4
, pp. 
381
-
392
, doi: .
Argote
,
L.
(
1999
),
Organizational Learning: Creating, Retaining & Transferring Knowledge
,
Kluwer
,
Norwell, MA
.
Armstrong
,
J.S.
and
Overton
,
T.S.
(
1977
), “
Estimating non response bias in mail surveys
”,
Journal of Marketing Research
, Vol. 
14
No. 
3
, pp. 
396
-
402
, doi: .
Arthur
,
W.B.
(
2007
), “
The structure of invention
”,
Research Policy
, Vol. 
36
No. 
2
, pp. 
274
-
287
, doi: .
Atuahene-Gima
,
K.
and
Murray
,
J.Y.
(
2007
), “
Exploratory and exploitative learning in new product development: a social capital perspective on new technology ventures in China
”,
Journal of International Marketing
, Vol. 
15
No. 
2
, pp. 
1
-
29
, doi: .
Baldwin
,
C.Y.
and
von Hippel
,
E.
(
2011
), “
Modeling a paradigm shift: From producer innovation to user and open collaborative innovation
”,
Organization Science
, Vol. 
22
No. 
6
, pp.
1399
-
1417
, doi: .
Beinhocker
,
E.D.
(
1999
), “
Robust adaptive strategies
”,
MIT Sloan Management Review
, Vol. 
40
No. 
3
, pp. 
95
-
106
.
Beltagui
,
A.
,
Rosli
,
A.
and
Candi
,
M.
(
2020
), “
Exaptation in a digital innovation ecosystem: the disruptive impacts of 3D printing
”,
Research Policy
, Vol. 
49
No. 
1
, 103833, doi: .
Berchicci
,
L.
(
2013
), “
Towards an open R&D system: internal R&D investment, external knowledge acquisition and innovative performance
”,
Research Policy
, Vol. 
42
No. 
1
, pp. 
117
-
127
.
Borgh
,
M.V.D.
,
Cloodt
,
M.
and
Romme
,
A.G.L.
(
2012
), “
Value creation by knowledge‐based ecosystems: evidence from a field study
”,
R&D Management
, Vol. 
42
No. 
2
, pp. 
150
-
169
, doi: .
Bruno
,
G.
,
Taurino
,
T.
and
Villa
,
A.
(
2018
), “
An approach to support SMEs in manufacturing knowledge organization
”,
Journal of Intelligent Manufacturing
, Vol. 
29
No. 
6
, pp. 
1379
-
1392
, doi: .
Burgelman
,
R.A.
and
Grove
,
A.S.
(
2007
), “
Let chaos reign, then rein in chaos--repeatedly: managing strategic dynamics for corporate longevity
”,
Strategic Management Journal
, Vol. 
28
No. 
10
, pp. 
965
-
979
, doi: .
Caner
,
T.
and
Tyler
,
B.B.
(
2015
), “
The effects of knowledge depth and scope on the relationship between R&D alliances and new product development
”,
Journal of Product Innovation Management
, Vol. 
32
No. 
5
, pp. 
808
-
824
.
Cao
,
L.
(
2017
),
Research on Intellectual Capital, Dual Learning and Product Innovation Performance in Horizontal Alliances
,
Nanjing University of Technology
,
Nanjing
.
Cao
,
Y.
,
Zhou
,
R.
,
Zhou
,
H.
and
Akira
,
N.
(
2019
), “
Research on the relationship between resource aggregation, dual learning, and enterprise innovation performance
”,
Science of Science and Management of S.& T.
, Vol. 
40
No. 
6
, pp. 
94
-
106
,
(in Chinese)
.
Caputo
,
F.
(
2024
), “
Human–artificial intellectual capital…beyond a fragmented perspective
”,
Journal of Intellectual Capital
, Vol. 
25
Nos
5-6
, pp. 
1026
-
1041
, doi: .
Capaldo
,
A.
,
Lavie
,
D.
and
Petruzzelli
,
M.A.
(
2017
), “
Knowledge maturity and the scientific value of innovations
”,
Journal of Management
, Vol. 
43
No. 
2
, pp. 
503
-
533
, doi: .
Carnabuci
,
G.
and
Operti
,
E.
(
2013
), “
Where do firms' recombinant capabilities come from? Intraorganizational networks, knowledge, and firms' ability to innovate through technological recombination
”,
Strategic Management Journal
, Vol. 
34
No. 
13
, pp. 
1591
-
1613
, doi: .
Cattani
,
G.
(
2005
), “
Preadaptation, firm heterogeneity, and technological performance: a study on the evolution of fiber optics, 1970-1995
”,
Organization Science
, Vol. 
16
No. 
6
, pp. 
563
-
580
, doi: .
Cattani
,
G.
and
Malerba
,
F.
(
2021
), “
Evolutionary approaches to innovation, the firm, and the dynamics of industries
”,
Strategy Science
, Vol. 
6
No. 
4
, pp. 
265
-
289
, doi: .
Chen
,
D.N.
,
Liang
,
T.P.
and
Lin
,
B.
(
2005
), “
An ecological model for organizational knowledge management
”,
Data Processor for Better Business Education
, Vol. 
50
No. 
3
, pp. 
11
-
22
, doi: .
Cheng
,
H.
,
Xu
,
X.
,
Li
,
Z.
and
Zhang
,
Z.
(
2024
), “
The synergy of inventor cooperative network dual embeddedness and firm innovation: the mediating role of ambidextrous learning
”,
Technology Analysis and Strategic Management
, Vol. 
36
No. 
1
, pp. 
89
-
104
, doi: .
Chesbrough
,
H.
(
2004
), “
Managing open innovation
”,
Research-Technology Management
, Vol. 
47
No. 
1
, pp. 
23
-
26
, doi: .
Chesbrough
,
H.
and
Chen
,
L.E.
(
2013
), “
Recovering abandoned compounds through expanded external IP licensing
”,
California Management Review
, Vol. 
55
No. 
4
, pp. 
83
-
101
, doi: .
Chesbrough
,
H.
and
Crowther
,
A.K.
(
2010
), “
Beyond high tech: early adopters of open innovation in other industries
”,
R & D Management
, Vol. 
36
No. 
3
, pp. 
229
-
236
, doi: .
Chi
,
M.
,
Wang
,
J.
and
Wang
,
W.
(
2022
), “
Research on the influence mechanism of enterprise innovation performance under the background of digital transformation-based on the mixed method of NCA and SEM
”,
Studies in Science of Science
, Vol. 
40
No. 
2
, pp. 
319
-
331
,
(in Chinese)
.
Choi
,
H.J.
,
Ahn
,
J.C.
,
Jung
,
S.H.
and
Kim
,
J.H.
(
2020
), “
Communities of practice and knowledge management systems: effects on knowledge management activities and innovation performance
”,
Knowledge Management Research and Practice
, Vol. 
18
No. 
1
, pp. 
53
-
68
, doi: .
Chung
,
H.
,
Yang
,
Z.
and
Huang
,
P.H.
(
2015
), “
How does organizational learning matter in strategic business performance? The contingency role of guanxi networking
”,
Journal of Business Research
, Vol. 
68
No. 
6
, pp. 
1216
-
1224
, doi: .
Chung
,
D.
,
Cho
,
T.S.
and
Kang
,
J.
(
2018
), “
The linkage between tmt knowledge diversity and firm-level innovation: the role of organisational search scope and managerial discretion
”,
International Journal of Technology Management
, Vol. 
78
No. 
3
, pp. 
208
-
233
, doi: .
Churchill
,
G.A.
and JrJr
(
1979
), “
A paradigm for developing better measures of marketing constructs
”,
Journal of Marketing Research
, Vol. 
16
No. 
1
, pp. 
64
-
73
, doi: .
Codini
,
A.
,
Abbate
,
T.
and
Petruzzelli
,
A.M.
(
2022
), “
Business model innovation and exaptation: a new way of innovating in SMEs
”,
Technovation
, Vol. 
119
No. 
5
, 102548.
Colombelli
,
A.
,
Krafftb
,
J.
and
Quatraro
,
F.
(
2013
), “
Properties of knowledge base and firm survival: evidence from a sample of French manufacturing firms
”,
Technological Forecasting and Social Change
, Vol. 
80
No. 
8
, pp. 
1469
-
1483
, doi: .
D'Este
,
P.
(
2005
), “
How do firms' knowledge bases affect intra-industry heterogeneity? An analysis of the Spanish pharmaceutical industry
”,
Research Policy
, Vol. 
34
No. 
1
, pp. 
33
-
45
.
Dew
,
N.
and
Sarasvathy
,
S.D.
(
2016
), “
Editor's choice: exaptation and niche construction: behavioral insights for an evolutionary theory
”,
Industrial and Corporate Change
, Vol. 
25
No. 
1
, pp. 
167
-
179
, doi: .
Dew
,
N.
,
Sarasvathy
,
S.D.
and
Venkataraman
,
S.
(
2004
), “
The economic implication of exaptation
”,
Journal of Evolutionary Economics
, Vol. 
14
No. 
1
, pp. 
69
-
84
, doi: .
Ding
,
S.
,
Li
,
W.
and
Xu
,
L.
(
2020
), “
Research on the relationship between IT capability and enterprise innovation performance——based on the perspective of dual learning regulaition
”,
Science and Technology Progress and Policy
, Vol. 
37
No. 
10
, pp. 
90
-
98
,
(in Chinese)
.
Duan
,
C.
and
Gu
,
Y.
(
2021
), “
The mechanism of new product architecture innovation based on ambidextrous learning and organizational unlearning: a case study of work circle, a SME management cloud platform
”,
Journal of Management
, Vol. 
18
No. 
11
, pp.
1589
-
1599
,
[In Chinese]
.
Eisenhardt
,
K.M.
and
Martin
,
F.J.A.
(
2000
), “
Dynamic capabilities: what are they?
”,
Strategic Management Journal
, Vol. 
21
No. 
10
, pp. 
1105
-
1121
, doi: .
Feng
,
W.
and
Chen
,
H.
(
2022
), “
The effects of ambidextrous innovation on organizational resilience of high-tech enterprises: the moderating role of knowledge scope and knowledge balanc
”,
Science of Science and Management of S.& T.
, Vol. 
43
No. 
4
, pp. 
117
-
135
,
(in Chinese)
.
Filippini
,
R.
,
Güttel
,
W.H.
and
Nosella
,
A.
(
2012
), “
Ambidexterity and the evolution of knowledge management initiatives
”,
Journal of Business Research
, Vol. 
65
No. 
3
, pp. 
317
-
324
, doi: .
Ganzaroli
,
A.
,
De Noni
,
I.
and
Pilotti
,
L.
(
2014
), “
The role of social entrepreneurship in leveraging exaptation in locked-in industrial districts: the case of idrogenet in the industrial district of lumezzane
”,
Innovation: The European Journal of Social Science Research
, Vol. 
27
No. 
3
, pp. 
254
-
274
, doi: .
Garud
,
R.
,
Gehman
,
J.
and
Giuliani
,
A.P.
(
2016
), “
Technological exaptation: a narrative approach
”,
Industrial and Corporate Change
, Vol. 
25
No. 
1
, pp. 
149
-
166
, doi: .
Garud
,
R.
,
Gehman
,
J.
and
Giuliani
,
A.P.
(
2018
), “
Serendipity arrangements for exapting science-based innovations
”,
Academy of Management Perspectives
, Vol. 
32
No. 
1
, pp. 
125
-
140
, doi: .
Ghemawat
,
P.
,
Ricart
,
J.E.
and
Costa
,
I.
(
1993
), “
The organizational tension between static and dynamic efficiency
”,
Strategic Management Journal
, Vol. 
14
No. 
S2
, pp. 
59
-
73
, doi: .
Gould
,
S.J.
and
Vrba
,
E.S.
(
1982
), “
Exaptation: a missing term in the science of form
”,
Paleobiology
, Vol. 
8
No. 
1
, pp. 
4
-
15
, doi: .
Gregory
,
T.R.
(
2008
), “
The evolution of complex organs
”,
Evolution Education and Outreach
, Vol. 
1
No. 
4
, pp. 
358
-
389
, doi: .
Hambrick
,
D.C.
and
Mason
,
P.A.
(
1984
), “
Upper echelons: the organization as a reflection of its top managers
”,
Academy of Management Review
, Vol. 
9
No. 
2
, pp. 
193
-
206
, doi: .
Harrigan
,
P.
,
Miles
,
M.P.
,
Fang
,
Y.
and
Roy
,
S.K.
(
2020
), “
The role of social media in the engagement and information processes of social CRM
”,
International Journal of Information Management
, Vol. 
54
No. 
54
, 102151, doi: .
Haunschild
,
P.R.
and
Sullivan
,
B.N.
(
2002
), “
Learning from complexity: effects of prior accidents and incidents on airlines' learning
”,
Administrative Science Quarterly
, Vol. 
47
No. 
4
, pp. 
609
-
643
, doi: .
Huang
,
Z.S.
,
Lu
,
Y.J.
,
Chau
,
Y.K.
and
Zeng
,
H.
(
2020
), “
Influence of ambidextrous learning on eco-innovation performance of startups: moderating effect of top management's environmental awareness
”,
Frontiers in Psychology
, Vol. 
11
, 111976, doi: .
Huber
,
G.P.
(
1991
), “
Organizational learning: the contributing processes and the literatures
”,
Organization Science
, Vol. 
2
No. 
1
, pp. 
88
-
115
, doi: .
Huston
,
L.
and
Sakkab
,
N.
(
2006
), “
Connect and develop
”,
Harvard Business Review
, Vol. 
84
No. 
3
, pp. 
58
-
66
.
Jiménez
,
L.D.
,
Dittmar
,
C.E.
and
Portillo
,
V.P.J.
(
2021
), “
Cooperation and relationship in the triple helix model of innovation
”,
International Journal of Knowledge and Learning
, Vol. 
14
No. 
1
, pp. 
1
-
9
, doi: .
Jin
,
L.
,
Zhao
,
Y.
and
Yin
,
X.
(
2025
), “
How do small and medium-sized enterprises specializing in manufacturing industry drive technological innovation through scenes: a longitudinal case study of Ebonord from 2010 to 2024
”,
China Soft Science
, No. 
5
, pp. 
41
-
54
.
Katila
,
R.
and
Ahuja
,
G.
(
2002
), “
Something old, something new: a longitudinal study of search behavior and new product introduction
”,
Academy of Management Journal
, Vol. 
45
No. 
6
, pp. 
1183
-
1194
, doi: .
Kati
,
J.
,
Argyro
,
A.
and
Paavo
,
R.
(
2018
), “
Organization of knowledge ecosystems: prefigurative and partial forms
”,
Research Policy
, Vol. 
47
No. 
8
, pp. 
1523
-
1537
.
Katkalo
,
V.S.
,
Pitelis
,
C.N.
and
Teece
,
D.J.
(
2010
), “
Introduction: on the nature and scope of dynamic capabilities
”,
Industrial and Corporate Change
, Vol. 
19
No. 
4
, pp. 
1175
-
1186
, doi: .
Kim
,
N.
and
Atuahene-Gima
,
K.
(
2010
), “
Using exploratory and exploitative market learning for new product development
”,
Journal of Product Innovation Management
, Vol. 
27
No. 
4
, pp. 
519
-
536
, doi: .
Lee
,
V.
,
Choong
,
C.
,
Wong
,
K.
and
Ooi
,
K.
(
2015
), “
Organisational learning as a mediating factor of TQM practices and technological innovation: an empirical analysis of Malaysia's manufacturing firms
”,
Int. J. of Innovation and Learning
, Vol. 
18
No. 
2
, pp. 
222
-
249
, doi: .
Levinthal
,
D.A.
and
March
,
J.G.
(
1993
), “
The myopia of learning
”,
Strategic Management Journal
, Vol. 
14
No. 
2
, pp. 
95
-
112
, doi: .
Li
,
F.
,
Xu
,
Y.
,
Du
,
P.
and
Xu
,
J.
(
2021a
), “
The influence of alliance management ability on subversive technological innovation of enterprises in the era of digital economy-the mediating role of knowledge flow and the regulating role of knowledge reconstruction ability
”,
Science and Technology Progress and Policy
, Vol. 
39
No. 
4
, pp. 
80
-
90
,
(in Chinese)
.
Li
,
T.
,
Ma
,
J.
and
Li
,
Y.
(
2021b
), “
Attach importance to drug research and development based on exaptation and strengthen technical support for major epidemic prevention and control
”,
Forum on Science and Technology in China
, No. 
1
, pp. 
127
-
135
,
(in Chinese)
.
Li
,
Z.
,
Sun
,
K.
and
Zhao
,
J.
(
2021c
), “
How can enterprise knowledge base regulate the performance of multi-source knowledge acquisition——threshold effect based on knowledge depth and breadth
”,
Studies in Science of Science
, Vol. 
39
No. 
2
, pp. 
303
-
312
,
(in Chinese)
.
Li
,
T.
,
Ma
,
J.
and
Dong
,
X.
(
2022
), “
Integrated technology exaptation in ecosystem context
”,
Studies in Science of Science
, Vol. 
40
No. 
8
, pp. 
1452
-
1461
.
Li
,
X.
,
Peng
,
Z.
and
Li
,
K.
(
2024
), “
Impact of boundary-spanning search on firm innovation performance: a strategic knowledge integration perspective
”,
Journal of Knowledge Management
, Vol. 
28
No. 
10
, pp. 
3075
-
3103
, doi: .
Lin
,
C.
,
Yu
,
C.
and
Wu
,
D.
(
2015
), “
Research on the impact of exploratory learning and utilitarian learning on disruptive innovation in enterprises
”,
R&D Management
, Vol. 
27
No. 
6
, pp. 
19
-
28
.
Liu
,
W.
,
Beltagui
,
A.
and
Ye
,
S.
(
2021
), “
Accelerated innovation through repurposing: exaptation of design and manufacturing in response to COVID‐19
”,
R&D Management
, Vol. 
51
No. 
4
, pp. 
410
-
426
, doi: .
Liu
,
G.
,
Wang
,
W.
,
Duan
,
Y.
,
Chin
,
T.
and
Mirone
,
F.
(
2025
), “
How entrepreneurial orientation influences innovation performance? The effect of knowledge coupling
”,
Journal of Knowledge Management
, Vol. 
29
No. 
1
, pp. 
247
-
258
, doi: .
Lumpkin
,
G.
and
Lichtenstein
,
B.
(
2005
), “
The role of organizational learning in the opportunity‐recognition process
”,
Entrepreneurship: Theory and Practice
, Vol. 
29
No. 
4
, pp. 
451
-
472
, doi: .
Luo
,
Z.
(
2009
), “
Research on the development of organizational technology innovation ability based on knowledge attribute
”,
Modern Management Science
, Vol. 
9
No. 
3
, pp. 
53
-
55
,
(in Chinese)
.
Luo
,
Z.
,
Callaert
,
J.
,
Zeng
,
D.
and
Looy
,
B.V.
(
2024
), “
Knowledge recombination, environmental turbulence and firms' innovation quality: the evidence from Chinese pharmaceutical industry
”,
European Journal of Innovation Management
, Vol. 
27
No. 
1
, pp. 
69
-
95
, doi: .
Lv
,
C.
,
Peng
,
C.
and
Yang
,
H.
(
2021
), “
Dual learning, improvisation ability and sustainable development performance-the moderating role of manager relationship
”,
Contemporary Economic Management
, Vol. 
43
No. 
9
, pp. 
40
-
49
,
(in Chinese)
.
Lv
,
C.
,
Peng
,
C.
and
Cao
,
D.
(
2023
), “
Dual learning, innovation driven process, and digital transformation: the moderating effect of digital capabilities
”,
Journal of Systems Management
, Vol. 
32
No. 
2
, pp. 
379
-
394
,
(in Chinese)
.
Ma
,
J.
and
Li
,
T.
(
2022
), “
Science-based innovation failure rescue: expanding the adaptive perspective
”,
Science and Technology Progress and Policy
, Vol. 
39
No. 
16
, pp. 
29
-
38
,
(in Chinese)
.
March
,
J.G.
(
1991
), “
Exploration and exploitation in organizational learning
”,
Organization Science
, Vol. 
2
No. 
1
, pp. 
71
-
87
, doi: .
Marinelli
,
L.
,
Crupi
,
A.
,
Sarto
,
D.N.
and
Dominique
,
L.
(
2024
), “
Unveiling knowledge ecosystem dimensions for MSMEs' digital transformation, toward a location-based brokerage
”,
Technovation
, Vol. 
136
, 103086, doi: .
Marquis
,
C.
and
Huang
,
Z.
(
2010
), “
Acquisitions as exaptation: the legacy of founding institutions in the U.S. commercial banking industry
”,
Academy of Management Journal
, Vol. 
56
No. 
6
, pp. 
1441
-
1473
, doi: .
Mastrogiorgio
,
M.
and
Gilsing
,
V.
(
2016
), “
Innovation through exaptation and its determinants: the role of technological complexity, analogy making and patent scope
”,
Research Policy
, Vol. 
45
No. 
7
, pp. 
1419
-
1435
.
Mccarthy
,
I.P.
,
Collard
,
M.
and
Johnson
,
M.
(
2017
), “
Adaptive organizational resilience: an evolutionary perspective
”,
Current Opinion in Environmental Sustainability
, Vol. 
28
No. 
10
, pp. 
33
-
40
, doi: .
Men
,
B.
and
Lan
,
H.
(
2023
), “
Closed network and enterprise innovation performance: the moderating effect of knowledge base and knowledge diversity
”,
Frontiers of Science and Technology of Engineering Management
, Vol. 
42
No. 
5
, pp. 
67
-
74
,
(in Chinese)
.
Mokyr
,
J.
(
2000
),
Evolutionary Phenomena in Technoloe. Gical Change. Ziman, J. Technological Innovation as an Evolutionary Process
,
Cambridge University Press
,
Cambridge
.
Nie
,
L.
,
Gong
,
H.
and
Lai
,
X.
(
2024
), “
Awakening ‘sleeping patents’: time delay, frequency, and key core technology innovation in knowledge recombination
”,
Nankai Business Review
, Vol. 
27
No. 
8
, pp. 
86
-
97+160
.
Osepchuk
,
J.M.
(
1984
), “
A history of microwave heating applications
”,
IEEE Transactions on Microwave Theory and Techniques
, Vol. 
32
No. 
9
, pp. 
1200
-
1224
, doi: .
Omezzine
,
F.
and
Bodas Freitas
,
I.M.
(
2022
), “
New market creation through exaptation: the role of the founding team's prior professional experience
”,
Research Policy
, Vol. 
51
No. 
5
, 104494, doi: .
Perry-Smith
,
J.E.
and
Mannucci
,
P.V.
(
2017
), “
From creativity to innovation: the social network drivers of the four phases of the idea journey
”,
Academy of Management Review
, Vol. 
42
No. 
1
, pp. 
53
-
79
, doi: .
Podsakoff
,
P.M.
,
MacKenzie
,
S.B.
,
Lee
,
J.Y.
and
Podsakoff
,
N.P.
(
2003
), “
Common method biases in behavioral research: a critical review of the literature and recommended remedies
”,
Journal of Applied Psychology
, Vol. 
88
No. 
5
, pp. 
879
-
903
, doi: .
Polese
,
F.
,
Carrubbo
,
L.
,
Caputo
,
F.
and
Sarno
,
D.
(
2018
), “
Managing healthcare service ecosystems: abstracting a sustainability-based view from hospitalization at home (HaH) practices
”,
Sustainability
, Vol. 
10
No. 
11
, p.
3951
, doi: .
Por
,
G.
and
Molloy
,
J.
(
2000
), “
Nurturing systemic wisdom through knowledge ecology
”,
The systems thinker
, Vol. 
11
No. 
8
, pp. 
1
-
5
.
Ren
,
S.
,
Gao
,
T.
and
Xu
,
H.
(
2019
), “
Exaptation as a source of innovation: a literature review
”,
Foreign Economics and Management
, Vol. 
41
No. 
1
, pp. 
17
-
30
,
(in Chinese)
.
Rong
,
X.
,
Yang
,
Z.
and
Sun
,
Y.
(
2024
), “
Inventors' brokerages dynamic and exploratory innovation: the moderating role of knowledge diversity
”,
Scientometrics
, Vol. 
129
No. 
5
, pp. 
2533
-
2558
, doi: .
Sanz-Valle
,
R.
,
Naranjo-Valencia
,
J.C.
,
Jiménez-Jiménez
,
D.
and
Perez-Caballero
,
L.
(
2011
), “
Linking organizational learning with technical innovation and organizational culture
”,
Journal of Knowledge Management
, Vol. 
15
No. 
6
, pp. 
997
-
1015
, doi: .
Savino
,
T.
,
Messeni
,
P.A.
and
Albino
,
V.
(
2017
), “
Search and recombination process to innovate: a review of the empirical evidence and a research agenda
”,
International Journal of Management Reviews
, Vol. 
19
No. 
1
, pp. 
54
-
75
, doi: .
Sedita
,
S.R.
,
Blasi
,
S.
and
Ganzaroli
,
A.
(
2022
), “
Exaptive innovation in constraint-based environments: lessons from COVID-19 crisis
”,
European Journal of Innovation Management
, Vol. 
25
No. 
6
, pp. 
549
-
566
, doi: .
Shen
,
H.
,
Sun
,
J.
,
Huang
,
C.
and
Mao
,
H.
(
2020
), “
Geographical scope, research and development mode, and innovation value acquisition of knowledge sources
”,
Studies in Science of Science
, Vol. 
38
No. 
7
, pp. 
1285
-
1293
,
(in Chinese)
.
Shoaib
,
A.B.
and
Kehinde
,
M.
(
2019
), “
The diversity of knowledge sources and its impact on firm-level innovation
”,
European Journal of Innovation Management
, Vol. 
22
No. 
4
, pp. 
681
-
714
.
Soto-Acosta
,
P.
,
Popa
,
S.
and
Martinez-Conesa
,
I.
(
2018
), “
Information technology, knowledge management and environmental dynamism as drivers of innovation ambidexterity: a study in SMEs
”,
Journal of Knowledge Management
, Vol. 
22
No. 
4
, pp. 
824
-
849
, doi: .
Sun
,
J.
,
Liu
,
L.
and
Chen
,
L.
(
2019
), “
Visual analysis of the research progress and Frontier evolution of knowledge dynamic ability
”,
Science and Technology Progress and Policy
, Vol. 
36
,
14
, pp. 
151
-
160
,
(in Chinese)
Tang
,
C.
,
Shi
,
Y.
and
Cai
,
R.
(
2022a
), “
Correlations of resource bricolage and exaptation with low-cost breakthrough innovations: moderating effect of organizational agility
”,
Frontiers in Psychology
, Vol. 
13
, 846629, doi: .
Tang
,
C.
,
Shi
,
Y.
,
Zhu
,
X.
and
Li
,
Y.
(
2022b
), “
How exaptation mediates the effects of customer participation on low-cost innovation: the role of strategic flexibility
”,
Frontiers in Psychology
, Vol. 
13
, 10165244, doi: .
Tho
,
N.D.
and
Duc
,
L.A.
(
2021
), “
Team psychological capital and innovation: the mediating of team exploratory and exploitative learning
”,
Journal of Knowledge Management
, Vol. 
25
No. 
7
, pp. 
1745
-
1759
, doi: .
Toothaker
,
L.E.
(
2017
), “
Multiple regression: testing and interpreting interactions
”,
Journal of the Operational Research Society
, Vol. 
45
No. 
1
, pp. 
119
-
120
, doi: .
Wanaswa
,
P.S.
,
Awino
,
Z.B.
,
Ogutu
,
M.
and
Owino
,
J.
(
2021
), “
Technological innovation and competitive advantage: empirical evidence from large telecommunication firms
”,
International Journal of Business and Management
, Vol. 
16
No. 
10
, p.
21
, doi: .
Wang
,
L.
and
Zhao
,
L.
(
2023
), “
Research on the mechanism of patent pool's influence on exaptation in digital economy
”,
Science Research Management
, Vol. 
44
No. 
5
, pp. 
166
-
173
,
(in Chinese)
.
Wang
,
Y.
,
Xie
,
W.
and
Lan
,
H.
(
2012
), “
IT capability and strategic flexibility: the mediating effects of exploratory and exploitative learning, reconstruction and sustainable enterprise innovation
”,
Soft Science
, Vol. 
34
No. 
9
, pp. 
85
-
89
,
(in Chinese)
.
Wang
,
W.
,
Cui
,
W.
,
Sun
,
X.
and
Tang
,
X.
(
2017
), “
The influence of knowledge scope and indirect connection on the creativity of key developers
”,
Frontiers of Science and Technology of Engineering Management
, Vol. 
36
No. 
5
, pp. 
36
-
42
,
(in Chinese)
.
Wang
,
H.
,
Zeng
,
D.
and
Chen
,
P.
(
2020a
), “
The impact of enterprise knowledge restructuring on technological innovation performance: the moderating effect of knowledge base relationship characteristics
”,
Nankai Business Review
, Vol. 
23
No. 
1
, pp. 
53
-
61
,
(in Chinese)
.
Wang
,
J.
,
Cao
,
N.
and
Ye
,
M.
(
2020b
), “
Multidimensional knowledge search, knowledge reconstruction and sustainable enterprise innovation
”,
Soft Science
, Vol. 
34
No. 
9
, pp. 
85
-
89
,
(in Chinese)
.
Wang
,
X.
,
Han
,
X.
and
Li
,
H.
(
2023
), “
The impact of innovation openness on innovation performance of manufacturing corporates——moderating effect based on knowledge reorganization
”,
Finance Research Letters
, Vol. 
36
,
PA
, 103917, doi: .
Wang
,
J.
,
Liu
,
X.
and
Ren
,
X.
(
2024a
), “
The impact of dual learning and organizational resilience on enterprise technological innovation
”,
Science and Technology Management Research
, Vol. 
44
No. 
8
, pp. 
1
-
11
,
(in Chinese)
.
Wang
,
W.
,
Yin
,
X.
,
Coles
,
R.
and
Chen
,
J.
(
2024b
), “
More knowledge, better innovation? Role of knowledge breadth and depth
”,
Management Decision
, Vol. 
62
No. 
5
, pp. 
1576
-
1597
, doi: .
Weible
,
D.
(
2013
), “
Approaching a semiotics of exaptation: at the intersection between biological evolution and technological development
”,
Sign Systems Studies
, Vol. 
41
No. 
4
, pp. 
504
-
527
, doi: .
Xu
,
L.
,
Zeng
,
D.
and
Zhang
,
Y.
(
2018
), “
Research on the relationship between knowledge network density and dual innovation performance: the moderation effect based on the diversity of knowledge foundation
”,
R&D Management
, Vol. 
30
No. 
1
, pp. 
72
-
80
,
(in Chinese)
.
Yang
,
X.
(
2024
), “
Could ambidextrous learning improve employees’ disruptive innovation behavior? A double-mediation model
”,
Social Behavior and Personality
, Vol. 
52
No. 
9
, pp.
1
-
11
, doi: .
Yang
,
Z.
,
Xu
,
R.
and
Chen
,
J.
(
2022
), “
Building a knowledge ecosystem for digital innovation
”,
Tsinghua Business Review
, No. 
12
, pp. 
64
-
73
,
(in Chinese)
.
Yayavaram
,
S.
and
Ahuja
,
G.
(
2008
), “
Decomposability in knowledge structures and its impact on the usefulness of inventions and knowledge-base malleability
”,
Administrative Science Quarterly
, Vol. 
53
No. 
2
, pp. 
333
-
362
, doi: .
Ye
,
J.
,
Ren
,
H.
and
Hao
,
B.
(
2016
), “
The internal mechanism of the effect of external knowledge heterogeneity on the innovation performance curve: the perspective of knowledge reconstruction and absorptive capacity
”,
Science Research Management
, Vol. 
37
No. 
8
, pp. 
8
-
17
,
(in Chinese)
.
Yi
,
Y.
,
Gu
,
M.
,
Liu
,
Y.
and
Ma
,
R.
(
2018
), “
Dynamic capability, dual learning and new product development performance
”,
Science Research Management
, Vol. 
39
No. 
1
, pp. 
74
-
82
,
(in Chinese)
.
Yu
,
L.
and
Yu
,
B.
(
2024
), “
Research on the impact mechanism of knowledge restructuring on innovation performance of high tech enterprises
”,
R&D Management
, Vol. 
36
No. 
3
, pp. 
70
-
83
,
(in Chinese)
.
Yuan
,
X.
(
2009
), “
Formation mechanism and prevention of sleeping patent
”,
Science Research Management
, Vol. 
30
No. 
4
, pp. 
168
-
174
.
Zeng
,
D.M.
,
Zeng
,
Y.L.
and
Zou
,
S.M.
(
2015
), “
Effect of knowledge acquisition strategy on corporate performance: the moderating role of knowledge base
”,
Science of Science Management of S.& T.
, Vol. 
36
No. 
9
, pp. 
31
-
39
,
(in Chinese)
.
Zhang
,
B.
,
Yin
,
X.
and
Wu
,
P.
(
2024
), “
The impact of knowledge sharing on technology innovation in the strategic industry: an empirical study from the knowledge-based view
”,
Journal of the Knowledge Economy
, Vol. 
16
No. 
3
, pp. 
1
-
30
,
(prepublish)
, doi: .
Zhong
,
C.
,
Huang
,
R.
,
Duan
,
Y.
,
Sunguo
,
T.
and
Dello
,
S.A.
(
2024
), “
Exploring the impacts of knowledge recombination on firms' breakthrough innovation: the moderating effect of environmental dynamism
”,
Journal of Knowledge Management
, Vol. 
28
No. 
3
, pp. 
698
-
723
, doi: .
Zhu
,
P.
,
Miao
,
X.
and
Jin
,
S.
(
2024
), “
Boundary-spanning search and business model innovation: the joint moderating effects of innovative cognitive imprinting and environmental dynamics
”,
European Journal of Innovation Management
, Vol. 
27
No. 
1
, pp. 
310
-
333
, doi: .
Duan
,
Y.
,
Liu
,
S.
,
Mu
,
C.
,
Liu
,
X.
,
Cheng
,
E.
and
Liu
,
Y.
(
2023
), “
The moderating effect of managerial discretion on cross-border knowledge search and the innovation quality of high-tech firms in a global health emergency: evidence from China
”,
Journal of Knowledge Management
, Vol. 
28
No. 
1
, pp. 
121
-
155
, doi: .
Ju
,
D.
and
Shen
,
B.
(
2015
), “
Library as knowledge ecosystem
”,
Library Management
, Vol. 
36
Nos
4/5
, pp. 
329
-
339
, doi: .
Wang
,
H.
(
2017
), “
Does technological complexity inhibit the impact of dual learning on innovation performance
”,
Science and Technology Progress and Policy
, Vol. 
34
No. 
16
, pp. 
140
-
147
,
(in Chinese)
.
Published by Emerald Publishing Limited. This article is published under the Creative Commons Attribution (CC BY 4.0) licence. Anyone may reproduce, distribute, translate and create derivative works of this article (for both commercial and non-commercial purposes), subject to full attribution to the original publication and authors. The full terms of this licence may be seen at Link to the terms of the CC BY 4.0 licence.

or Create an Account

Close Modal
Close Modal