I treat textual data appearing on websites, such as discourse or descriptors, as qualitative, meaning it is information that is difficult to measure, count, or express in numerical terms. Descriptors provide key insight into the scope of members included within a tribal group website as well as the level of specificity members use to refer to themselves within the group. Fuzzy logic is a logical system dealing with the concept of partial truth with values ranging between completely true and completely false. This study uses fuzzy logic to associate specific numerical values with particular websites to illustrate levels of relational connectivity. The following description details the processes developed in the course of this study to measure levels of relational connectivity between websites, determine degrees of “fuzziness,” and construct network representations.

This analytical portion of the study remains experimental, relying upon subjective assumptions regarding the role of referential tribe-based descriptors within observed websites. Values are derived from a scale based upon referential specificity, ranging from 0.0 to 1.0. For instance, discourse and descriptors within the Amonsoquath Tribe of Cherokee website (Amonsoquath Tribe, No Date), which indicate its members identify as part of a specific tribe within the greater Cherokee nation context, received a value of 0.9. Conversely, the NativeWeb (NativeWeb, 2006) website, which often uses the indistinct “Indigenous” descriptor to refer to its members, was assigned a value of 0.2 due to its wide scope and ambiguity. The Caribbean Amerindian Certrelink website (CAC, 2006) denotes a slightly more specific “Amerindian” descriptor and geographic locale, receiving a value of 0.4. The following table represents the process used to determine the relational connectivity between these three Native group websites.

The relational connectivity number, or RCN, represents the degree of connectivity between nodes in a given dataset, based upon the level of specific ethnic distinction within each tribal site. A high RCN indicates that nodes exhibit a high level of tribal specificity, while a low RCN indicates broader pan-Indian qualities exist between elements in the dataset. The RCN is determined by dividing the sum of values of the dataset by the dataset’s cardinality (total number of element within the set). As shown above, the RCN for the data set of {A,B,C} is 0.5. The RCN value of 0.5 indicates that there is not a strong relation existing between this dataset, reflecting the broader, pan-Indian scope of the NativeWeb and Centrelink sites compared to the tribal specific Amonsoquath website. Once the initial RCN is found for a dataset, additional RCNs may be found for the subsets in order to display the relational bonds across edges within the dataset. The RCN may be determined for multiple datasets in order to display the level of connective bonds existing between much larger groups of data as long as the numerical values within the scale of referential specificity remains the same for all of the datasets.

Measuring the degree of “fuzziness” within a dataset is a powerful tool to analyze relationships existing between every node, represent points of comparison, and visualize the relational network. Based upon the concept of referential specificity, I determined the quantitative values for each of the fifteen websites originally cataloged and documented, which are displayed in the chart above. Substituting these values within the m1 and m2 variables, I applied the following equation to every relationship existing within the dataset.

In contrast to the RCN formula, which measures the relational connectivity within the entire dataset, this equation measures the degree of relational difference between two websites. Websites exhibiting similar RS values return an equally similar value and are grouped together. For example, the Biloxi-Choctaw (0.9) and Tanio-Tribe (0.8) return a value of 0.9 indicating tribal specific websites. Likewise, groups such as NativeWeb (0.2) and Aboriginal Connections (0.3) also return a value of 0.9, indicating a strong degree of similarity in regards to exhibition of ambiguous supratribal descriptors. A diagonal matrix representing the degree of relation between every website then forms when the fuzziness measurement formula is applied to the entire dataset (see Appendix). The diagonal matrix is introduced into Pajek, an open source network visualization application, where graphical representations of the websites’ relational structure are rendered. The following two representations offer markedly different perspectives on the relational network.

> 0.9 Threshold

In this instance the website network is rendered with a threshold of greater than 0.9, clearly illustrating the strongest relational connections—i.e. least degree of fuzziness—between nodes by severely limiting the number of edges. Beginning with NativeWeb, the inverted arc’s nodes increase in referential specificity as it descends, concluding with a pentagram-esque sub-network representing websites with the highest RS values. The arc’s declension represents a symbolic migration from broader pan-Indian to tribal specific identity. This construction symbolically illustrates the diversity of scope, referential descriptors, and ethnic distinction exhibited by websites encountered in this study.

No Thresholds placed on Network

This representation displays every possible relational connection between websites in the dataset. The immense amount of edges within this matrix of connective links mirrors the exponentially complex relational structure of websites on the Internet. I began this portion of my project aiming to represent purely qualitative data in a quantifiable format, in hopes of revealing meaningful patterns that previously would have remained undetected. Although the visualizations accurately and abstractly illustrate this study’s data, I was unable to arrive at a definitive method to truly produce new knowledge. Still, the process itself proved invaluable to my study, providing an alternate paradigm to consider how these groups are related rhetorically and structurally. The visualizations are physical representations of the intangible communicative space that is the Internet. This functionality alone provides insight into the formation of a broader community structure within a virtual context. Similarly, the varying degrees of relation between nodes in these representations present a framework to view pan-Indianism’s emergence online. Although all tribal websites are not directly related, they remain connected in some sense, providing a context for participants to identify with and ascribe to representations on multiple domains. This area of the study remains experimental, however the cognitive processes it required profoundly impacted the project as a whole, providing a meaningful contextual lens to view collected data and key concepts central to this study.