The notion of dominant terminology in bibliometric research

Drawing on the innovation literature (Utterback & Abernathy, 1975), we recall that for innovations or new designs, often a dominant design appears over time. In the first phase following the introduction of an innovative product, different technologies compete with each other for customers. Later, a dominant design – possibly with variants – emerges. A dominant design is a single architecture that establishes dominance in a product category (Abernathy & Utterback, 1978). A dominant design wins the allegiance of the marketplace, the one to which competitors and innovators must adhere if they hope to command a significant market following (Utterback, 1994). A design is called dominant if it acquires more than 50% of the market share of the product category and maintains this market share for four consecutive years (Anderson & Tuchman, 1990). Although at the origin of the idea, Utterback and Abernathy did not use the term “dominant design”, in their 1975 paper but referred to “dominant strategy” and “dominant type of innovations” instead.

Researchers studying that typical aspect of innovation have focused on two major questions: Will a dominant design emerge in a given product category, and, if it does, how long will it take for a dominant design to emerge (Srinivasan et al., 2006). Industry experts have investigated diverse sectors and have identified dominant designs in diverse product categories such as typewriters, automobiles, televisions, picture tubes, and nuclear power plants (Utterback, 1994). Dominant designs can occur at the level of the product, but also at the subsystems level, or the component level of a product: transistors, microprocessors, or configuration of cylinders in engines (Murmann & Frenken, 2006).

A dominant design often emerges from competition among different designs and is not necessarily the best. In the home video recorder market, the VHS of JVC gained market share against the superior Betamax recorder (Srinivasan et al., 2006). To avoid this competition between designs Philips offered its Compact Disk technology to Sony, based on two dozen codified standards, on which the sector has agreed (Peek, 2010). This strategy allowed to diminish investments in product development and to increase the global market share.

Another obvious example to illustrate that the dominant design is not always the best is the QWERTY keyboard on a computer. Whatever the original reason was to introduce QWERTY (Stamp, 2013), this reason was certainly not valid anymore for a computer keyboard. Of course, the fact that millions of users of the Latin alphabet (or should we say English speaking, as the French-speaking use AZERTY) were used to this layout was the main reason for keeping it also for computer keyboards.

However, not all product categories lead to a dominant design, and in many cases, several competing designs continue to co-exist (Srinivasan et al., 2006). Well-known examples are the PC with Apple and Microsoft operating systems (Hagedoorn, Carayannis & Alexander, 2001), and the cell phones with Android versus Apple (Cecere et al., 2015).

The same phenomenon, but related to the use of terms, seems to have taken place in the sciences. An example of the expanding applications of medical printing has been investigated by Chepelev et al. (2017). They found out that different medical disciplines have worked in rather isolated silos and have described their research activity in multiple subdomain-specific isolated terminologies. The analysis of their publications has identified several self-isolating research clusters. Three major terms have been used: “3D Printing”, “Rapid Prototyping” and “Additive Manufacturing”. Yet, since 2015 “3D Printing” has become the dominant term, except for dentistry and otolaryngology which use “rapid prototyping”.

A similar phenomenon occurs in the social sciences. For instance, the variety of labels and terminologies around LGBTQ+ in academic research. Thelwall et al. (2021) studied the use of 74 LGBTQ+ terms in Scopus. They found out that the term “homosexuality” which has dominated early research, has nearly disappeared for other terms such as “gay”, “lesbian” or “bisexual”. As of now, there is no dominant terminology; the authors suggest different research needs are reflected in a plurality of terminology.

Both cases illustrate the need for a common ontology and universal precise terminology to assist in information retrieval and comparison, knowledge integration, and collaboration (Chepelev et al., 2017; Thelwall et al., 2022). A recent case concerns the research on Covid-19, the fastest-expanding research ever. While the first publications mentioned the generic term “coronavirus”, very soon researchers replaced it with the specific term “Covid-19” which became the dominant term in only a few months (Fassin, 2021).

A linguistic study on the use of dominant words and the organization of a shared vocabulary has been presented in PNAS (Pagel et al., 2019). The researchers apply the Neo-Darwinist theory of natural or random drift of generic variants to the common use of words. However, Pagel et al. (2019) suggest that directional selection can to some degree move people to adopt new lexical variants. When many competing new alternative words are proposed, no external authority directs word use, but speakers spontaneously self-organize a shared vocabulary. The same phenomenon of positive frequency-dependent selection arises in specialized research networks. Our study on terminology in bibliometrics illustrates this occurrence (Fassin & Rousseau, 2023).

Another important portion of the dissemination of innovation terms consists of fads, the rapid spread of novel cultural forms and ideas (Pagel et al., 2019). This notion of fads has been analyzed in management concepts (Abrahamson et al., 1996); fads – or at least new terms – are often launched by management consultants to promote the specific solutions that they propose.

Not all innovations lead to a dominant design. In the same way, given the wax and wane of innovative notions, not all new competing terms lead to a unique dominant terminology. For closely related but different themes different dominant terms may continue to co-exist. Especially new sub-themes in a certain field may want to distinguish themselves from the major domain.

In some domains, different closely related themes in the same domain may continue to be used despite overlapping. An international comparative study on the perception of themes related to business ethics such as corporate social responsibility, sustainability, and corporate governance shows how the perception of the meaning differs, and how these differences vary in different languages (Fassin et al., 2015). It also illustrates how the closely related notions of corporate social responsibility and sustainability have grown together to the overarching term corporate responsibility. Nevertheless, researchers from different management fields continue to use their own preferential terminology.

We now apply the issue to the broad field of bibliometrics/science of science. Different terms proposed for a newly developed subfield compete in the first phase. Although many terms may be proposed, most do not reach high dissemination.

A typical example is the field we nowadays usually refer to as webometrics. This term was introduced by our Danish colleagues Almind and Ingwersen in 1997 (Almind & Ingwersen, 1997). Competing terms such as netometrics (Bossy, 1995), webometry (Abraham, 1996; 1997), web bibliometry (Chakrabarti et al., 2002), and cybermetrics, introduced by Aguillo in 1997 as the title of an e-journal, did not make it. Yet, also the term ‘webmetrics’ is not often used. Webmetrics could be considered a better term because the connecting oh-sound is only there to facilitate pronunciation (this is also the case for scientometrics) or to sound like bibliometrics. More importantly, ‘webo’ sounds like the Spanish word for eggs, which is used as a vulgar term for testicles. One can imagine that Spanish-speaking colleagues definitely prefer the term webmetrics.

Of course, we all know how the original term ‘statistical bibliography’ has been replaced (in two steps) by the term ‘bibliometrics’ (Otlet, 1934; Pritchard, 1968), with ‘librametrics’ and even ‘informetrics’ as losing contenders. We see the term ‘informetrics’ as the better term as it is not tied to books and libraries, and indicates the relation to information in its generality.

The recently introduced term ‘altmetrics’ (Priem et al., 2010) has also been scrutinized as certainly not the best. According to Rousseau and Ye (2013), it is just a new way of doing informetrics, it sounds like “old metrics”, and certainly other alternatives will emerge. Nevertheless, it looks like it is a term with staying power. A study that focuses on social media citations may still prefer the term altmetrics over bibliometrics, also as a keyword, for reasons of precision and differentiation. The choice between a general domain and a sub-domain remains a sensitive issue for many researchers. Ideally is to integrate both in the keyword list.

Recent investigations by us (Fassin & Rousseau, 2023; Rousseau & Fassin, 2022) showed that the term “bibliometrics” has emerged as a dominant term, while scientometrics survived as the second choice (mainly because of its use as a journal title). Yet, we expect an increase in the use of the terms altmetrics, and science of science.

Although we will not go into this in detail, we note that for software used in bibliometric and network studies, it seems that VOSviewer (van Eck & Waltman, 2007; 2010) is nowadays the dominant one, but with several other contenders, such as Pajek (Batagelj & Mrvar, 1998; de Nooy et al., 2005) and CiteSpace (Chen, 2003).

In analogy to the dominant design in product development, we introduced the idea of dominant terminology in the information sciences and provided several examples in the domain of science of science.

This phenomenon illustrates the dilemma between a plurality of terminologies to reflect different research needs and the need for a common ontology to improve information retrieval and knowledge integration. The increasing interdisciplinarity of contemporary science has enlarged this dilemma.