ECMI Minorities Blog. Could a new level of digitalisation offer minority language media outlets a strategy to attract a wider audience?
Author: Craig Willis | https://doi.org/10.53779/SMPX1996
When digitalisation is mentioned in the context of traditional media, it is often in relation to the challenges or threats such outlets face – for example, increased competition and reduced revenue opportunities. This is the same in the field of minority language media yet with the added existential dimension; most minority language spheres only have one or two outlets in a given language. Such fears are not new and have been discussed across the last 10 years and more; for example concern raised in the Welsh context about audience migrating to English-language media and a feeling that technology was working against minority languages. Moreover, concern has previously been expressed that machine translation was unreliable and error-prone in cases of Irish, Scottish Gaelic, and Welsh. However, technology does not stand still and thus the process of digitalisation is a dynamic evolving one – is it time to take stock and reconsider whether minority language media can benefit from recent technological advancements and grow its audience, or indeed whether smaller minority languages without media outlets could now enter the media sphere?
Technological developments, such as a move away from terrestrial television into digital, have had mixed effects on television; whilst it allows more broadcasting space, it also thus brings further competition. A 2019 report written for the European Charter for Regional and Minority Languages’ Committee of Experts (COMEX) highlighted this as a specific issue for minority language broadcasting. In a similar vein, minority language newspapers have become increasingly focused on their online content and this again brings opportunities of wider audience but also threats of more competition. Indeed there has been a shift towards further online news outlets against a decline of physical newspapers, something also confirmed by the COMEX report which points to changing reading habits as a cause. The question therefore arises as to how MLM outlets can sustain a large-enough audience and whether a strategy of audience expansion to non-speakers could be part of that.
In the realm of minority language broadcasting, the Scottish Gaelic channel ‘BBC Alba’ operates a different model to many of its comparable cases, in that a significant part of their strategy is to provide content for the wider audience and not just the core Gaelic community. This is a reflection of the demographic situation whereby only around 1.5% of Scotland speaks Gaelic, and thus a channel could only be approved in the knowledge that it would have a broad appeal. Whilst this situation may not provide the ideal service for the core Gaelic-speaking group, it does enable the channel to be sustained and offers exposure to Gaelic to a wide audience, therefore helping to increase the visibility and status of the language in Scotland. It has therefore shown that it is possible to attract a wide non-speaking audience by creating unique or exclusive content embedded in the local context, such as sports, music and arts, and documentaries, facilitated by accessibility in the form of subtitling. Other channels such as S4C in Wales, TV3 in Catalonia, ETB1 in the Basque Country and TVG in Galicia do state linguistic normalisation and reaching a wide audience within their statutes and/or aims; yet for the most part, they focus almost entirely on the speaking audience. Moreover, BBC Alba was launched in 2008 into an already digitalised media landscape, and as such was immediately available on-demand through the BBC iPlayer and utilised many social media channels to promote programming. Thus, the BBC Alba model is relatively unique in minority language broadcasting and may offer a strategy which others could follow, even in the sphere of newspapers – as the following paragraph unpacks.
One such element would be the enhancement of auto-translation in recent years, which includes many minority languages now available through services such as Google Translate. Furthermore, the most recent additions to Google Translate have been developed through ‘zero-shot machine translation’, allowing only original language texts to be used, rather than requiring pairs of languages. This approach can therefore be useful for “improving accessibility for languages that are considered to be under-resourced”. Moreover, the automation of live subtitling has also advanced significantly in recent years, to allow for much more accurate real-time captioning. This, combined with auto-translation, could improve the service ML television could offer, without the high costs of human transcription and translation. For existing channels, this could assist with widening their appeal beyond the speaking group, by offering more extensive subtitling or written website content using auto-translation – an aspect which may help channels under threat from falling viewing figures. Similarly, it could allow advertisers to subtitle in a minority language or even be encouraged to create minority language adverts in the knowledge that they can be simply translated through the combination of automated subtitling and auto-translation. Moreover, such technologies could also form part of a proposal for minority languages that do not yet have a full channel but are campaigning for one – for example, the calls for a Cornish broadcaster in the UK – and allow the proposal to demonstrate a model similar to BBC Alba.
In the area of newspapers, taking the example of the Basque language, auto-translation allows one to read, for example, the Basque language newspaper, Berria, in Spanish or English (or indeed any other language supported by Google Translate). Logically, this could be done directly in the browser for the online version of the newspaper, but physical newspapers could also be translated through a smartphone and the Google Translate app. Whilst this would obviously be a tedious way to read an entire newspaper, newspapers such as Gara, which publishes mostly in Spanish but with some Basque content, could be read through an app by non-Basque speakers for the parts of the newspaper which they cannot understand. These are obviously not services offered or promoted by either Berria or Gara, but they could feasibly fully offer their websites in different languages through auto-translation. Such a technological tool could then be used by an MLM outlet such as Berria to target a wider audience alongside the core group, who may be attracted by the local content they cannot find elsewhere but have previously been unable to engage due to the linguistic barrier. Similarly, with newspapers like Gara, which currently publish in a mixture of majority and minority language, high quality auto-translation could allow such publications to amend this balance and concentrate on more minority language content first. Finally, members of the minority diaspora further afield may be able to engage with digital media content from their ‘motherland’ even if they do not have the full language skills – for example, the Basque diaspora in the Americas. Across all of these examples, it should also be borne in mind that such flexibility would also benefit learners of the language, who would be able to switch between languages at their own leisure when necessary.
Even without auto-translation, the digitalisation of newspaper content has seen growth in readership numbers in comparison to physical newspaper sales, a growth enhanced by the COVID-19 pandemic. The case of the German newspaper in Denmark, Der Nordschleswiger, is one such case which is now reaching around 12,000 daily unique visits, in comparison to subscriber numbers of just over 1,000 before its daily print edition was wound down in February 2021. Moreover, due to its digital form, it has been able to attract readers from further afield, including both Danish speakers in Denmark and German speakers in Germany (interview with Gwyn Nissen) – it is not known whether the former are using translation tools to change the content from German to Danish. Similar findings were observed in the Swedish media of Finland, with an increasing audience interest from Sweden.
However, it is worth considering what sacrifices this change in strategy could have for the core-audience. In the case of BBC Alba, one of the major aspects of its strategy is to include non-optional English subtitles in almost all programming. Moreover, its most popular content is often designed around a non-speaking audience – for example its sports coverage does not contain half-time analysis and instead shows mini documentaries, often with a lot of interviews in English. This is accepted as part of a strategy but if ML broadcasters switched to such a format, its existing viewer base might not find this acceptable. In the case of digital newspapers, Der Nordschleswiger began to change the way it writes in the knowledge that the readership is increasingly drawn from beyond the community – at least regarding certain content. Chief Editor Gwyn Nissen remarked that their content would often need to provide more background detail than previously, for example listing affiliations of persons mentioned or local organisations – things obvious to a member of the community but not to the wider audience. Similarly, outlets need to ensure that they continue to broadcast/publish in the minority language and not create content in the majority language because it may be easier or cheaper, as this would defeat the objective of MLM and remove the creative and professional use of the minority language in the domain of media – as well as the production industry such outlets support. Thus, the challenge with considering a wider audience is to maintain the service provided to the core audience whilst offering a wider appeal.
To conclude, the current phase of digitalisation evidently warrants further scrutiny as well as a consideration that much of the academic literature on ‘digitalisation’ in the context of minority language media has been written before the recent developments discussed above. Indeed, such advances in machine translation are very much still theoretical for all but the major global languages – to this end, Google have stated their intention to make their translation as good for minority languages as it is for hegemonic languages like Spanish or German. Thus, the likely further advances will improve upon past issues of mistranslations, particularly through the enhancement of zero-shot development. Given the existing engagement from majority audiences with MLM, it is likely that audiences will utilise such tools at their own initiative. The opportunity for MLM is to guide this process to their advantage and potentially grow their audience beyond the speaker group, helping to showcase the value of a ML and its surrounding culture to the broader society – as the model of BBC Alba has been shown so successfully to do.
***
This blog post was prepared by the author in their personal capacity. The views expressed in this blog post are the sole responsibility of the author concerned and do not reflect the view of the European Centre for Minority Issues.