Introduction

This analysis uses network visualization to examine relationships between 111 influential 1890s-1990s designers. The study reveals natural communities that transcend conventional categorizations, showing how design ideas flowed across regions and styles.

RESEARCH QUESTION

Did specific geographical regions exhibit more substantial influence within particular design categories, and how might this relate to recognition levels for designers from those regions?

Methods

The dataset comprises 111 designers from 17 countries, 62 schools, and seven art movements, from the same dataset for the Map Visualization, but enriched with nodes. Designer attributes include gender, style, country, and school affiliation. Connections between designers represent relationships including shared schools and spousal connections, with weighted edges (1-3) indicating relationship strength.

We used Gephi’s ForceAtlas2 layout with modularity detection to create a graph showing node connections. Based on earlier iterations, this graph has a density of 20.2% and a modularity percentage of 31.6%. The size of each node indicates its influence, which is determined by the number of connections it has. We made several visualizations, including:  

• Nodes colored by geographical region  
• Nodes sized by weighted degree  
• Nodes colored by design style  

REFERENCES

I will create a static poster for my network visualization to learn more about the concept. However, once I learn more about CSS and JS, I look forward to revisiting the visualization and making it more interactive.

DATASET

The dataset combines information on 111 chair designers with their professional and personal connections. Initial data collection focused on basic designer attributes (gender, style, country, school) and relationships between them (spouse, shared_school).

This raw data was processed in Gephi to generate network statistics. The software calculated various metrics, including degree measurements, weighted values, and centrality measures. The most significant enrichment came through Gephi’s modularity algorithm, which identified four natural communities among the designers with a modularity percentage of 31.6% and overall graph density of 20.2%.

The final enhanced dataset reveals the interconnections between 111 designers from 17 countries, representing 62 schools and seven art movements, clustered into four distinct clusters.

VISUALIZATION

To understand the underlying factors driving the Modularity Class and its visualization, I created multiple visualization approaches using ForceAtlas 2:

FINAL RESULTS

This network analysis provides valuable insights into the ecosystem of mid-century chair design, revealing how communities formed not just around geography or stated design philosophies, but through actual working relationships and influence patterns. Here is the full overview of the main pattern and information.

Please explore the information for each modularity cluster to learn more.

Five key patterns emerged from the visualization:

1. Design movements transcend geography while maintaining regional characteristics
2. Community structures vary significantly from highly cohesive (American) to more loosely connected (European Functionalist)
3. Bridge figures played crucial roles in transmitting ideas across different design approaches
4. Regional design centers (Milan, Copenhagen, New York) formed distinctive creative ecosystems

We can also break the network analysis to reveal four distinct modularity groups with unique characteristics

KEY FINDINGS

Modernist Core Dominance Reflects Historical Power. The American Modernist cluster (orange) comprises 40.5% of the network with the highest connectivity density (0.65). This demonstrates how mid-century design was dominated by American interpretations of modernism through commercial partnerships with corporations like Herman Miller and Knoll.

Regional Style Adaptations Maintain Distinct Identities. Despite sharing modernist principles, each modularity group developed regionally distinctive approaches: American pragmatism, Scandinavian organic functionalism, Italian experimentation, and European theoretical foundations, while operating within a Western-defined framework.

Educational Pathways Shape Network Clusters. Key educational institutions (Cranbrook, Royal Danish Academy, Bauhaus, Ulm School) strongly influenced the formation of design communities, with graduates often clustering together regardless of nationality.

Bridge Figures Facilitate Cross-Regional Exchange. A few designers (approximately 7-9% of the network) account for roughly 15% of cross-modularity connections, serving as crucial conduits for design ideas to travel between different communities.

Varying Community Structures Reveal Different Design Ecosystems. The density gradient across modularity groups (from 0.65 to 0.35) reveals fundamentally different organizational approaches – from the tightly collaborative American system to the more outward-facing European functionalist approach.

---

Reflection

BIASES & LIMITATIONS

• The dataset may over-represent well-documented designers and relationships, potentially obscuring less prominent but still significant figures.
• Might be biased towards the range of weight regarding the relationships due to self-labeling through research on Wikipedia
• Connections don’t distinguish between different types of relationships (collaboration, influence, mentorship)
• The static representation masks the temporal evolution of these relationships, which developed over several decades.

FUTURE DIRECTIONS

Enhancing Data and Context

• Develop a more nuanced coding schema for relationships between designers to differentiate types of connections.
• Create interactive visualizations using Sigma.JS with pop-up windows containing designer information and significant works.
• Implement timeline/geomapping functionality to show the evolution of connections over the mid-century chair design.

Advanced Analytical Approaches

• Create ego networks for key bridge figures to better understand their specific role in connecting different design communities.
• Apply additional network metrics to identify less obvious but significant patterns of influence.
• Incorporate additional metadata about materials, techniques, and clients to reveal deeper patterns.
• Expanding Scope and Addressing Bias
• Integrate social and historical context to better understand forces shaping the geographical and stylistic landscape of 20th-century design.