The changing norms and standards of scholarly journal articles. A response to Pietilä’s “Peoples Conceptions of the Mass Media”

1 Introduction

Reading decades-old journal articles is fascinating as it reveals how much the norms and standards of scholarly journal articles have changed. Pietilä’s “People’s Conceptions of the Mass Media” (1976) is a good example. If this article had been submitted today, it would in all likelihood have been desk rejected. This is, however, not to say that it does not have merits.

2 The problems

The overall purpose of this article is to investigate how people’s conceptions of the mass media change when a new mass medium – television – is introduced. Methodologically, it draws upon surveys among people in Finland. So far so good, one might think. However, the problems are manifold.

To begin with, no research problem is presented. When I say research problem, I refer to the problem or the puzzle that has not been solved by previous theory and research and that is of theoretical relevance and importance. Hence, it is not the same thing as a research gap. A research gap simply refers to something that has not been investigated before, but the fact that something has not been investigated before does not mean that it is relevant and interesting to investigate. For something to be scientifically interesting to investigate, it has to have some theoretical relevance, and both the research problem and the theoretical relevance need to be explicated.

Secondly, no theoretical framework is provided. In fact, Pietilä’s article is virtually devoid of any explicit theoretical reasoning, save for a few mentions of cognitive dissonance. Beyond that, not a single theory is mentioned, and there is not a single reference to other scholars and their work –not even Festinger’s (1957) pioneering study on cognitive dissonance. This is peculiar in itself, but also because the article discusses and investigates conceptions of the functions of different mass media. Thematically, it is thus close to uses and gratifications research, which at that time was under development with several important studies already in print (Katz, Blumler, and Gurevitch, 1973; Katz, Haas, and Gurevitch, 1973). Thus, there were theories that Pietilä could have built upon, referred to, and contributed to, but the article is written as if there was no relevant theory or as if theory was not relevant.

Third, the methodology and data are both opaque and inadequate given contemporary standards. As presented, the article builds on surveys in 1965, 1966 and 1967, and the major focus is on comparing the significance and functions ascribed to different media among those who at the time owned or did not own television sets, or alternatively bought a television set between the years. In places, Pietilä writes about “a panel research design”, suggesting that the same respondents participated in the different surveys. Yet, no information is provided about the target population, sample, response rate, representativeness, or even the number of responses in each survey. Only in the summary, placed after the main article and the endnotes, is it mentioned that the data “were collected by personal interviews in Finish Lapland before (in September 1966) and after (in September 1967) television programs could be watched there”, and that the “results are based on the answers of 443 respondents”. No information is given about the response rate, how many participated in the first and the second rounds of interviews, or about whether those 443 respondents participated in both. The composition of the final sample is also not described, nor the representativeness. Similarly, there is no information about the respondents in the 1965 study. This opaqueness and lack of information makes it impossible to assess the results and how far they can be generalized.

Fourth, it also has to be noted that the empirical analyses are highly descriptive and quite unsophisticated. Mainly, Pietilä is comparing the distribution of responses between different groups of respondents and between different years. There are no significance tests, no regressions or any other slightly more advanced analyses, although some comparisons of means are mentioned in the running text. However, no statistical data is provided. Beyond that, Pietilä also constructs some versatility indices, but again, no details are provided. In the tables, in most cases only percentages are presented, with no information about the number of respondents in each group. Altogether then, the analyses and presentation of the results fall short of what can be expected of scholarly articles, not only by contemporary standards.

3 The contributions

Based on this criticism, it might be assumed that this article offers no research contribution at all. A more generous reading, however, is possible if one disregards contemporary norms and standards of scientific journal articles. To begin with, the basic question of how the use and perceptions of different media change when new media are introduced is interesting. In light of the continuously shifting media environments and the emergence of new forms of media (cable tv, digital media, social media, political alternative or partisan media, mobile platforms, and who knows what might come next), it also continues to be relevant. This is evidenced by the many studies over the years that have investigated how the internet, social media, and more recently mobile platforms, for instance, influence people’s media use and the extent to which the use of newer media substitutes or complements the use of older forms of media (see, for example, Dutta-Bergman, 2004; Nelson and Lei, 2018; Skogerbø and Winsvold, 2011; Westlund and Färdigh, 2015). In that context, Pietilä’s article was pioneering.

Although Pietilä (1976) does not explicitly refer to uses and gratifications, his article also contributes to uses and gratifications research by investigating how important respondents think different media are with respect to news function, information function, art function and entertainment function, both in themselves but also relative to each other. This theory continues to be relevant in the 21^st century (Rubin, 2009; Ruggiero, 2000), and is typically applied whenever new types or forms of media appear, such as the internet (Kaye and Johnson, 2002; Papacharissi and Rubin, 2000) and more recently TikTok (Bucknell Bossen and Kottasz, 2020; Vaterlaus and Winter, 2021). This is not to deny that uses and gratifications research still suffers from several problems. Among these are studies that are often too compartmentalized and produce their separate typologies of motive, a lack of clarity regarding key concepts such as psychological motivations and needs, and that the notion of active, rational audiences is too simplistic and disregards how structural, social, situational and habitual factors influence media use beyond whatever motivations people might have (e.g., Ruggiero, 2000; Toff et al., 2024; Webster, 2014). This holds true also in contemporary digital media environments, in which media use is shaped not only by what Thorson and Wells (2016) label personal curation, but also social, journalistic, strategic and algorithmic curation. In other words, motivations, preferences and perceived functions matter (Prior, 2007; Strömbäck et al., 2013), but so do many other factors. It is also far from evident how much self-reports about motivations, preferences, perceived functions, and their self-perceived impact reveal about which factors in the end determine what media and media content people use and the role of sought gratifications in that context.

What I found most interesting in Pietilä (1976), however, is one of the findings (with the caveats regarding methodology and data noted earlier). As could be expected, one change following the acquirement of television that the study found was that the importance of radio suffered a severe decline. In 1966, before they had a TV, 64 percent of respondents stated that radio would be the medium that they would miss the most if it ceased to exist. In 1967, after they had acquired a TV set, the corresponding share was just 22 percent. Pietilä also notes that television in the process replaced radio as “a general, all-purpose medium” (p. 158). While that change is not very surprising, what is interesting is that “this loss is not all in favor of television” (Pietilä, 1976, p. 153). While 27 percent of TV-owners stated that TV would be the medium which they would miss the most if it ceased to exist, the medium that was deemed the most important was in fact newspapers, where the share stating that they would miss that medium the most went from 20 to 41 percent. In other words: “in 1967, the most important medium for television owners is the newspaper” (Pietilä, 1976, p. 153).

This finding is interesting for at least two reasons. First, it illustrates that newer media do not necessarily render older media less important or obsolete: rather they reconfigure the media environment and may even render some older media more important. This is noteworthy, since we still have printed newspapers despite frequent predictions that digital media will kill newspapers (Thottam, 1999). For sure, digitalization has undermined news media’s traditional business models, heavily decreased advertising revenue and undermined readership and the number of paid subscriptions (Newman et al., 2023; Nielsen and Ganter, 2022; Usher, 2021), but newspapers still exist, still reach large audiences, still produce most of the original news, and still influence the public and political agendas. Although some newspapers have gone under, others have been successful in adapting to an increasingly digital media world. Second, it illustrates the importance of avoiding technological determinism. This is no less important in contemporary media environments, where artificial intelligence (AI) currently is the hottest buzzword, and which is predicted to have major repercussions in virtually all domains. This is not to suggest that AI is not or will not be important or consequential. Rather, it is a reminder of what I think of as a rule of thumb that is relevant whenever new technologies appear: The effects of new technologies are usually overestimated in the shorter run but underestimated and misunderstood in the longer run. The underlying reason is quite simple: in the shorter run, the focus is on the technologies per se and their affordances, and new technologies often bring hype that leads people to hope for or fear major changes. In the longer run, the effect of any technology depends not only on their affordances, but rather how they meet and interact with cultural, social, political and economic needs, and those become evident only over time.

From that perspective, it would have been interesting if Pietelä had followed up his study a few years later, when TV had become more established and people more used to it. Maybe the results then would have been quite different.

4 Discussion and conclusion

I began this commentary with the critical note that Pietilä’s article (1976) would probably be desk rejected if it had been submitted today and judged by contemporary norms and standards of scholarly journal articles. This is, however, not unique for this article. One might, for example, wonder whether Noelle-Neuman’s article on the spiral of silence (1974), Gerbner’s article on cultivation theory (1970), Katz, Blumler, and Gurevitch’s (1973) article on uses and gratifications, or McCombs and Shaw’s (1972) article on agenda-setting would have been accepted for publication today.

The answer is probably no. Still, these and other articles were published, and over time they have proven to be of great scholarly value. This raises questions about whether the continuously higher and tougher theoretical, methodological and data analytical norms and standards have a potential downside.

In my mind, there is no doubt that the higher and tougher norms and standards have significantly increased the quality and value of empirical research. Ultimately, research is about getting as close as possible to generally valid knowledge about different phenomena and their causes and consequences, and that requires a solid theoretical foundation, the use of high-quality data, and the application of the most applicable analytical tools and techniques available at any given time. It also requires that the process is transparent. Otherwise, other scholars cannot assess the methodology, data, or results, and it becomes virtually impossible to replicate the studies. The higher and tougher norms are also essential in political and media environments where dis- and misinformation flourish, and where pseudo-scientific research is one tactic used to push certain narratives or undermine trust in established science (Oreskes and Conway, 2012; Vowles, 2024).

Still, more descriptive studies also have value, and sometimes it might be important and productive for scholars to present theoretical ideas that are not yet fully thought through, or where there is not yet sufficient or high-quality data to test out the ideas. Of course, it is possible in research seminars and at conferences, but the reach of those is limited, meaning that the ideas do not get the scrutiny and feedback that might help develop and flesh them out. For all their merits, the format and process of reviewing and publishing scholarly articles can have a preservative effect. While the solution is not to relax the contemporary norms and standards of scholarly articles, this is worth reflecting upon and discussing.