The study, published in Discover Education, examined mission and vision statements from higher education institutions in seven Iberoamerican countries: Argentina, Chile, Colombia, Costa Rica, Spain, Mexico, and Peru. The researchers, Rafael Fernández-Concha, Jubalt Álvarez-Salazar, and Percy Marquina Feldman, screened a sampling frame of 1,145 institutions and collected 901 mission statements and 798 vision statements publicly available on official university websites in June 2021. Because the frame was defined in advance and every institution in it was checked, the corpus amounts to a census of publicly disclosed identity statements within those systems rather than a selective sample.

What makes the study technically notable is that all analysis was performed on the original Spanish texts. The authors deliberately avoided translation, which can blur meaning and introduce drift, and instead processed the documents with the UDPipe toolkit for tokenization, part-of-speech tagging, and lemmatization — reducing words to their dictionary forms so that, for example, different inflections of calidad (quality) count as the same concept. Stopwords such as articles and prepositions were stripped away, and for each country the team extracted the fifty most frequent nouns, adjectives, and verbs, normalizing frequencies by corpus size to keep systems of different scale comparable.

From these cleaned inventories the researchers built semantic co-occurrence networks, in which nodes represent lemmatized nouns and adjectives and edges connect terms that appear within the same statement, using a sentence-bounded window of plus or minus five tokens. Verbs were excluded to reduce process-oriented noise and focus on stable conceptual associations. Edge weights reflected how often pairs of terms co-occurred, and networks were sparsified with a minimum co-occurrence threshold of two. The team then computed standard network indicators — size, average degree, density, degree centralization, average path length, and diameter — and identified structurally influential terms using Bonacich centrality, a measure that rewards being connected to other well-connected nodes. Hierarchical clustering on Pearson correlation matrices revealed how terms group into thematic families, and Quadratic Assignment Procedure correlations compared whole network structures across countries.

The headline finding is a shared legitimation lexicon. Eight terms consistently occupy the most central positions in mission-statement networks across all seven countries: universidad (university), formación (training), conocimiento (knowledge), sociedad (society), profesional (professional), investigación (research), calidad (quality), and institución (institution). A second tier — including desarrollo (development), académico (academic), educación (education), científico (scientific), and comunidad (community) — appears in at least four countries. In vision statements, a slightly smaller core emerges: university, quality, society, development, and knowledge. Cross-country correlations of the most frequent terms generally exceeded 0.70, indicating substantial lexical similarity and what the authors describe as a regional discourse community.

This convergence is exactly what institutional theory would predict. Universities operate in policy-dense environments shaped by accreditation agencies, quality-assurance frameworks, government regulation, and international rankings — pressures that reward institutions for signaling alignment with recognized expectations of quality, research, and social contribution. In the language of organizational sociology, universities adopt a common vocabulary as a rationalized signal of legitimacy, a way of becoming legible to regulators, accreditors, students, and funders. The study’s sensitivity checks reinforced the robustness of this core: pooled centrality profiles closely matched the unweighted mean of national profiles, and the main structural patterns held when the co-occurrence window, stopword lists, and sparsification thresholds were varied.

But the deeper story lies in the differentiation. Mexico presents the densest and most evenly connected semantic structures, with very low degree centralization — a cohesive, repetitive discursive core in which no single term dominates. Costa Rica sits at the opposite pole, with the lowest network density, longer average distances between concepts, and greater lexical dispersion, consistent with more diverse institutional self-presentation. Spain emphasizes academic and international orientations, elevating terms such as internacional and crítico. Chile highlights innovación (innovation) and vinculación (linkage to society), Peru foregrounds responsabilidad (responsibility), competitivo (competitive), and sostenible (sustainable), while Costa Rica’s distinctive vocabulary includes ambiente (environment), vida (life), and excelencia (excellence). Argentina emphasizes extensión, the tradition of university outreach to the wider community.

Crucially, the clustering analysis shows that even shared terms are wired differently in different national systems. The word conocimiento (knowledge) associates with necesidad (need) in Argentina, with generación and misión in Chile, with compromiso, cultura, and humano in Colombia, and with environmental and usage-related terms in Mexico. In Peru, seven of the eight most central terms cluster together at a 90 percent similarity threshold — the tightest thematic integration observed — whereas in Costa Rica key terms often fail to group even at relaxed thresholds. The authors interpret these patterns as national grammars: distinctive ways of reweighting and recombining a common vocabulary rather than abandoning it. Differentiation, in other words, happens within the shared lexicon, not outside it.

The authors are careful about what the evidence can and cannot support. The comparative results are descriptive and do not establish causality; the analysis documents patterned associations consistent with institutional isomorphism but does not identify the effects of particular regulatory instruments. The dataset is also imbalanced — Mexico contributes more than half of all statements — and it captures a single moment, June 2021, rather than a longitudinal trajectory. The exclusion of verbs, while improving thematic clarity, limits insight into action-oriented language. Future work, the authors suggest, could incorporate verbs and multi-word expressions, track the corpus over time, and link discursive patterns to funding, accreditation outcomes, or student perceptions.

The practical implications are nonetheless striking. Mission and vision statements are not mere branding; they are governance-facing texts through which universities translate identity into a form recognizable to external audiences. The study warns of a dual risk: excessive convergence produces formulaic, boilerplate identities that tell stakeholders nothing distinctive, while excessive dispersion can erode the shared legitimacy that makes institutions mutually intelligible. The authors propose semantic indicators — density and centralization among them — as early-warning tools for quality-assurance agencies and policymakers, helping them ask whether accreditation frameworks encourage meaningful differentiation or merely template compliance. For university leaders, the message is not to invent novel vocabularies but to ensure their statements genuinely express what the institution prioritizes, whom it serves, and how it differs from its peers. In an era when every university claims quality, research, and social contribution, the real signal may lie in how those familiar words are put together.

Subject of Research: Semantic network analysis of university mission and vision statements across Iberoamerican higher education systems

Article Title: Semantic convergence and national differentiation in university mission and vision statements across Iberoamerican countries

Article References: Fernández-Concha, R., Álvarez-Salazar, J., & Feldman, P. M. (2026). Semantic convergence and national differentiation in university mission and vision statements across Iberoamerican countries. Discover Education, 5(1), Article 1151. https://doi.org/10.1007/s44217-026-02280-5

Image Credits: AI Generated

DOI: 10.1007/s44217-026-02280-5

Keywords: higher education, mission statements, vision statements, semantic network analysis, institutional isomorphism, Iberoamerica, quality assurance, organizational identity, corpus linguistics, accreditation, university governance, text mining

Now, researchers Daniel Guzzo and Daniela C. A. Pigosso of the Technical University of Denmark have published, in the Journal of Industrial Ecology, the first systematic map of how specific circular economy business model patterns connect to the rebound mechanisms that undermine them. Their study, published as open access research, does something the field has long lacked: instead of cataloguing isolated case studies of rebound, it builds a semantic network that traces the causal pathways linking design choices in circular business models to the economic and behavioral chains that activate rebound effects. The result is both a diagnostic tool for companies and a sobering structural insight about the circular economy itself.

The core of the method is an interpretation-based semantic network analysis, a technique that maps relationships between concepts within a body of knowledge rather than merely counting how often they co-occur. The researchers began by constructing a causal ontology, a formal specification of the concepts, categories, and relationships that structure the domain, indicating how change propagates through it. In their framework, a circular economy business model pattern is performed by an actor, which stimulates a trigger or a driver; that trigger then activates a rebound mechanism. Triggers are the factors that set rebound mechanisms in motion, such as freed-up budget or new revenues, while drivers are regulators that amplify or dampen the ultimate effect, such as social values or the substitutability between consumption options.

To populate the network, the team analyzed the most comprehensive empirically supported catalog of circular economy business model patterns in manufacturing to date: 24 industry-generic patterns, from collaborative services to circular supplies, accompanied by 34 complementary pattern cards. Using qualitative content analysis, they dissected each pattern description to identify the changes it causes in production and consumption dynamics that could stimulate a trigger or driver. Each stimulation instance was characterized by the specific change involved, the trigger or driver it represented, the actor affected, and the type of evidence, whether explicitly stated in the text, interpretively inferred, or logically extrapolated from a similar context. On the other side of the ledger, the researchers deconstructed a state-of-the-art catalog of 26 rebound mechanisms into their constituent triggers, drivers, and actors, allowing both knowledge domains to be joined within a single causal structure.

The scale of the resulting network is striking. The analysis charted 117 distinct instances of trigger or driver stimulation within circular business model dynamics, alongside 55 instances of rebound mechanism activation. Fifteen of the 24 business model patterns connect to five or more distinct triggers and drivers, and crucially, this high connectivity transcends strategy types. Patterns focused on effectiveness, meaning enhanced value creation at constant resource flows, showed the highest average number of connections at 6.54, but efficiency patterns averaged 5.2 and even the single sufficiency pattern, which aims to reduce the need for value creation altogether, showed six. In other words, no strategic intent is immune: the potential for rebound appears to be an inherent structural property of circular business models, not an occasional side effect of poor design.

Perhaps the most revealing finding is what dominates the network. Economic and financial factors account for 60 percent of all identified connections between business model patterns and triggers, with 70 of 117 connections, and they activate the highest number of unique rebound mechanisms, 18 in total. The most frequent individual triggers are revenues and operational costs. The network also shows that the type of actor matters: the manufacturer or service provider and the company as a customer are the two most central nodes, with 33 and 31 outgoing connections respectively, and their links are overwhelmingly economic. Individuals, by contrast, show a more even distribution between economic factors and consumer choices, reflecting that companies are targeted with tangible financial benefits while consumers are offered intangible ones like flexibility and convenience, which are themselves well-known rebound triggers.

The map also exposes blind spots. Several enabling patterns, which reinforce other patterns rather than acting directly, do not appear in the network at all because their influence on triggers is never articulated in the source texts. Some triggers identified in the business model literature, such as space savings and moral hazard, have no corresponding mechanism in the rebound catalog, while crucial factors known to drive rebound, such as excess of supply, are absent from the business model descriptions. Not a single trigger stimulation was associated with socio-cultural influences, even though such factors have been linked to rebound elsewhere. The authors attribute this asymmetry to a deep dependence on a value-capture paradigm in business model research and on neoclassical economics in rebound research, producing what they call a form of rebound myopia, a limited capacity to even communicate about rebound beyond a narrow set of economic factors.

To demonstrate the practical value of the network, the researchers worked through a detailed use case: the collaborative services pattern, in which companies offer shared access to products through a platform. In a hypothetical business-to-consumer scenario, the service provider may generate additional revenues through platform fees, potentially activating output, factor substitution, and economies-of-scale rebound mechanisms. Individual renters may benefit from increased convenience, which can activate motivational consumption and motivational substitution, while reduced product access costs free up budget that may trigger further mechanisms. With these pathways laid out, a company can devise strategies to address the specific triggers before they are released, rather than discovering the rebound after the fact. The authors suggest even concrete monitoring approaches, such as tracking damage reports per use against historical averages to detect moral hazard in car-sharing fleets.

The researchers are careful about what the network can and cannot do. Although it employs quantitative measures such as connection counts, a node’s prominence indicates theoretically relevant associations, not actual magnitude or likelihood in a specific case; a single, less-connected pathway could still have a disproportionately large impact in context. The focus on generalized archetypes also blurs the nuances of real cases, since a trigger like decreased cost can unfold very differently depending on whether it affects a renter’s budget or a lender’s revenues. The framework is intended for proactive early-stage reflection, revealing the scope of potential systemic responses, and its authors point toward simulation-based approaches and dynamic modeling as natural complements for exploring temporal dynamics and contextual dependencies the map cannot capture.

The broader implications reach beyond circular business models. The authors argue that the same causal ontology and analytical procedures can be adapted to examine rebound in other sustainability interventions, including product design, socio-technical transitions, and policy tools, and that the lens can even be inverted to deliberately stimulate secondary benefits, which share similar causal structures. By systematically connecting constructs that have long been studied in isolation, the study challenges the proposition that rebound effects are not necessarily tied to specific circular strategies, and it pushes the field from classifying individual cases toward identifying patterns based on relational knowledge. For a discipline that has often leaned on a win-win narrative, the message is uncomfortable but constructive: the very benefits that make circular business models attractive, from reduced costs to new markets, are frequently the same triggers that activate rebound. Making those pathways visible, the researchers contend, is the first step toward designing them out, and toward circular interventions that finally deliver their full sustainability promise.

Subject of Research: Mapping the causal associations between circular economy business model patterns and rebound effects using semantic network analysis

Article Title: Bridging circular economy business models and rebound effects: Mapping systemic associations with semantic network analysis

Article References: Guzzo, D., & Pigosso, D. C. A. (2026). Bridging circular economy business models and rebound effects: Mapping systemic associations with semantic network analysis. Journal of Industrial Ecology, 30(4), 1629-1643. https://doi.org/10.1007/s44498-026-00110-3

Image Credits: AI Generated

DOI: 10.1007/s44498-026-00110-3

Keywords: circular economy, rebound effects, business models, semantic network analysis, sustainability, causal ontology, triggers and drivers, industrial ecology, sustainable consumption, manufacturing, rebound mechanisms, circular strategies