The position of Tibor Braun’s Œuvre:

Bibliographic journal coupling

In: The Multidimensional World of Tibor Braun,

at the occasion of his 75th birthday, March 2007


Loet Leydesdorff

Amsterdam School of Communications Research (ASCoR), University of Amsterdam

Kloveniersburgwal 48, 1012 CX  Amsterdam, The Netherlands ;



At the occasion of Tibor Braun’s 75th birthday, I introduce the method of bibliographic journal coupling for the analysis of the knowledge base of a document set; in this case, the set of 183 articles published by Tibor Braun. The method enables the user to visualize the knowledge base of author-based or institution-based sets in terms of the journals which are cited in the sets.



At previous occasions, I analyzed the semantic space spanned in terms of co-occurrences of words (co-words) by Braun’s—at that time 81—publications (Leydesdorff, 1992) and, five years later, of the 37 webpages which could be retrieved using the AltaVista search engine at the time (Leydesdorff, 2002). In this study, I complement these studies by analyzing the position of Tibor Braun’s œuvre using bibliographic coupling within the set of 183 documents authored by him. The bibliographic information about these documents was downloaded from the ISI Web-of-Science on  7 February 2007, using the search string “au = Braun T and ci = Budapest”.


Two papers are bibliographically coupled if they share a common reference (Kessler, 1963). Bibliographic coupling thus reverses co-citation analysis by asking the question about the internal citation structure of a document set (Garfield, 2001). This structure represents the knowledge base of a set (Garfield et al., 2003). This representation of the knowledge base can be refined by using the journal names in the references as the coupling agents. The technique enables us to visualize the historical knowledge base of a set, while bibliographic coupling itself reveals only the results of the coupling in the present.




The software for this analysis is freely available from my website at and, respectively. The first program can be used for bibliographic coupling itself and the second for the refinement proposed in this papers. Both programs use downloads from the ISI Web-of-Science as their input, and generate files which are in the format of Pajek. Pajek is a visualization program which is freely available for academic usage at (De Nooy et al., 2005).


The output files are co-occurrence matrices and cosine-normalized matrices (Leydesdorff & Vaughan, 2006). The cosine is equivalent as a similarity measure to the Pearson correlation coefficient except that this measure does not normalize to the arithmetic, but to the geometric mean (Jones & Furnas, 1987). This is convenient in the case of non-normal distributions (Ahlgren et al., 1993). Cosine values fit into a vector space which, for various reasons, reveals more about structure in the data than the raw (co-occurrence) data (Leydesdorff, 2007 and forthcoming; Salton & McGill, 1983).




a. bibliographic coupling

Figure 1 shows the results of bibliographic coupling among these 183 documents in which Tibor Braun is at least one of the co-authors. The bibliographic coupling is based on 2,221 references in these documents; 122 (co-)authors are involved.

Figure 1: 122 authors bibliographically coupled to Tibor Braun’s œuvre.


The figure informs us that Braun’s activities are not limited to Scientometrics. From the perspective of scientometrics as a specialty, only a group of scholars on the right side of the picture are co-authors in this domain. This result raises questions about Braun’s other collaborations. The names of the collaborators are not sufficiently informative for indicating these collaborations.


b. The knowledge base of Braun’s network


The shift to not using the 2,221 cited documents as units for the coupling among the 122 authors involved, but the journals cited within these document enables us to make the co-authorship patterns visible in terms of scientific specialties. The 2,221 references were published in 236 sources more than once. The cosine-normalized matrix of this set of 236 sources is visualized in Figure 2. Labels are suppressed because the figure would no longer be readable.


Figure 2: 236 journals (and other sources) cited by the 183 documents of the set (N = 2,221).


Only 18 of these cited journals are part of the k-core of the journal Scientometrics. This group is made visible in Figure 3.


Figure 3: the k-core of Scientometrics (n = 18)


Figure 3 illustrates again the relatively marginal position of scientometric work in Braun’s œuvre. In addition to Scientometrics, Braun also initiated the journal Fullerene Science and Technology in 1992. This journal was renamed into Fullerenes Nanotubes and Carbon Nanostructures in 2002.


Figure 4: 47 journals in the k-core of Fullerenes


The mere size of the cluster suggests that this set of journals is central to Braun’s œuvre.


Figure 5 completes this presentation by showing the labels for all journals.

Figure 5: labelled set of 236 sources of 2,221 citations in Braun’s œuvre.


c. Powerlaws


Since one can wonder whether Braun’s citations of journals follow a specific type of distribution, let me report on the following findings.


Figure 6: 21 journals with more than one percent betweenness centrality


Figure 6 first shows the initial part of the distribution of the 236 journals represented in Figures 2 to 5. The betweenness centrality of the 21 journals contributing more than one percent to this measure follow a powerlaw (Katz, 2000). However, the tale of the distribution (207 journals) follows an exponential curve (Figure 7).



Figure 7: 207 journals with more than zero betweenness centrality


In another context, Leydesdorff & Bensman (2006) found the opposite effect for aggregated journal-journal citations: in that case, the tale followed a powerlaw, while the initial part (“the hook”) did not (Price & Thelwall, 2005). In summary, this author-based set is organized differently from a journal set. While aggregated journal citation networks are self-organizing, an author set is organized hierarchically. For example, Braun is very much the centre of this set as being a co-author in all the publications. This is illustrated in Figure 8 using the measure of betweenness centrality.


Figure 8: Betweenness centrality of Tibor Braun’s œuvre among a set of coauthors.


The result of Figure 8 is an artefact of my representation. J




The above results show that your œuvre has become central to an interdisciplinary network. Congratulations with your 75th birthday!



Ahlgren, P., Jarneving, B., & Rousseau, R. (2003). Requirement for a Cocitation Similarity Measure, with Special Reference to Pearson's Correlation Coefficient. Journal of the American Society for Information Science and Technology, 54(6), 550-560.

De Nooy, W., Mrvar, A., & Batagelj, V. (2005). Exploratory Social Network Analysis with Pajek. New York: Cambridge University Press.

Garfield, E. (2001). From bibliographic coupling to co-citation analysis via algorithmic historio-bibliography. Speech delivered at Drexel University, Philadelphia, PA, November, 27.

Garfield, E., Pudovkin, A. I., & Istomin, V. S. (2003). Why do we need algorithmic historiography? Journal of the American Society for Information Science and Technology, 54(5), 400-412.

Jones, W. P., & Furnas, G. W. (1987). Pictures of Relevance: A Geometric Analysis of Similarity Measures. Journal of the American Society for Information Science, 36(6), 420-442.

Katz, J. S. (2000). Scale Independent Indicators and Research Evaluation. Science & Public Policy, 27(1), 23-36.

Kessler, M. M. (1963). Bibliographic coupling between scientific papers. American Documentation, 14, 10-25.

Leydesdorff, L. (1992). The Organization of the Semantic Space of an Author. (At the occasion of Tibor Braun's 60th birthday.). Psientometrics 60, 19-24; at

Leydesdorff, L. (2002). "The Globalization of an Author," a webpage <!loet.html> for the 70th birthday of Tibor Braun, the Editor of Scientometrics. at

Leydesdorff, L. (2007). Visualization of the Citation Impact Environments of Scientific Journals: An online mapping exercise. Journal of the American Society of Information Science and Technology, 58(1), 207-222.

Leydesdorff, L. (forthcoming). “Betweenness Centrality” as an Indictor of the “Interdisciplinarity” of Scientific Journals. Journal of the American Society for Information Science and Technology.

Leydesdorff, L., & Bensman, S. J. (2006). Classification and Powerlaws: The logarithmic transformation. Journal of the American Society for Information Science and Technology, 57(11), 1470-1486.

Leydesdorff, L., & Vaughan, L. (2006). Co-occurrence Matrices and their Applications in Information Science: Extending ACA to the Web Environment. Journal of the American Society for Information Science and Technology, 57(12), 1616-1628.

Price, E., & Thelwall, M. (2005). The clustering power of low frequency words in academic webs. Journal of the American Society for Information Science and Technology, 56(8), 883-888.

Salton, G., & McGill, M. J. (1983). Introduction to Modern Information Retrieval. Auckland, etc.: McGraw-Hill.