Carmen Dayrell, Basil Germond (Lancaster University) and Celine Germond-Duret (Liverpool John Moores University)

1. Introduction

5^th November 2021 was COP26’s Ocean Action Day. The UK Presidency of the conference stressed the need “to take ambitious steps towards ocean health and resilience” in order to contribute to “our fight against climate change”. Ocean sustainability is contingent to citizens’ awareness of “the benefits they receive from the marine environment” (DEFRA, 2021, p.4). However, the sea is at the bottom of the list when it comes to public perception of global environmental issues (Potts et al., 2016).

This study examines the representation of the sea in the UK press, focussing on articles published in 2020 and analysing both national (broadsheet and tabloids) and regional newspapers (see section 3 for details). The goal is to unravel the way the sea is represented in the British written press, which is the third source of information about the marine environment (DEFRA, 2021, p.23). More specifically, we seek to explore: (i) the extent to which the sea is represented in purely technical, economic and opportunistic terms as opposed to emotional and identity terms; and (ii) how the media representation of the sea can inform our understanding of citizens’ connection to the sea.

• Findings in brief

Finding 1:        The narrative frequently represents the sea in terms of economic opportunities: resources, profit and job creation.

Finding 2:        The ‘marine environment’ (and not the ‘sea’ itself) is represented as a natural resource that must be preserved, protected, especially in view of sustaining the economic benefits from the sea.

Finding 3:        Newspapers frequently stress the negative impacts of climate change on the sea.

Finding 4:        An emotional lexicon can only be found in relation to aesthetic considerations.

Finding 5:        A sense of place can be found in relation to the seashore/coastal locations.

• Data

The analysis is drawn from three separate datasets, or ‘corpora’.  The National Corpus, divided into The National Broadsheet and The National Tabloid Subcorpora, and The Regional Corpus. Table 1 provides the number of texts and words included in each corpus.

Corpus	Number of texts	Number of Words
National corpus	28,406	22,076,726
National Broadsheet subcorpus	20,099	17,651,577
National Tabloid subcorpus	8,307	4,425,169
Regional corpus	11,545	6,700,126

Table 1: Number of texts and total number of words comprising each corpus

Figures 1 to 3 show the selection of individual newspaper titles within each dataset, with the overall number of articles per newspaper title.

Articles were published between 1/1/2020 and 31/12/2020. All texts were collected from a news aggregator service (LexisNexis), considering the printed version of the newspapers in their weekday and Sunday versions. The collection of individual texts proceeded on the basis of specific terms: we searched for articles that contained either ‘sea(s)’ or ‘ocean(s)’ and considered any type of articles in which those words appeared. This means that, in addition to news reports, the corpora include other types of texts such as editorials and letters to the editor.

For the national newspapers specifically, we selected the national editions only, thus excluding the Irish, Scottish and Northern Ireland editions. Duplicates were removed from all corpora. However, for The Regional Corpus specifically, we kept articles that published across different newspapers (usually part of the same media group) given that they would reach different audiences.

1. Methods

We used analytical techniques associated with the field of Corpus Linguistics to study the dominant narratives in the national and regional newspapers, and within each considering all newspaper titles in aggregate. The corpora were processed using the software WordSmith Tools version 7 and the software package LancsBox.

To provide an overview of the most distinctive linguistic characteristics of each corpus, we carried out ‘keyword’ analyses. Keywords are words that are more frequent in a corpus of interest (known as the ‘study’ corpus) than they are in another corpus (known as the ‘reference corpus’), where the difference is statistically significant. They can be interpreted as reflecting the most distinctive concepts and themes in a particular corpus.

We carried out three separate comparisons. We first generated the keywords in the National Corpus using the Baby+ edition of the British National Corpus 2014 (BNC2014-baby) as the reference corpus. We then carried out a similar procedure in the Regional Corpus. These procedures identified words that were salient in the National and Regional Corpora respectively in relation to a general corpus of British English. We then generated the keywords in the Regional Corpus using the National Corpus as the reference corpus so that we could identify words that were prominent in the Regional but not in the National Corpus.

For the calculation of keywords, we focused on words that occurred with a minimum frequency of 100 occurrences in one million words, in at least 1% of the total number of texts in the study corpus. This was to ensure that the analysis focused on words that were overall relatively frequent and occurred across various texts. In terms of statistical tests, we combined a statistical test of significance (the log-likelihood test) with an effect size measure (Log-Ratio). The log-likelihood test tells us to what extent differences in frequencies between the two corpora (the study and the reference corpus) is statistically significant. It was applied considering a critical value higher than 6.63 (p < 0.01). Log-Ratio measures how big the difference is. The higher the Log-Ratio score, the larger the difference. The log-ratio calculation therefore gives us the words whose frequencies are proportionally higher in the study corpus. For the analysis, we focused on the 20 keywords with the highest Log-Ratio score in each calculation.

Keywords were interpreted by examining their ‘collocations’ through close reading of their ‘concordance lines’. Collocation analyses explore co-occurrence relationships between words, and therefore makes it possible to study the narratives or discourses that a word is part of. Concordance lines refer to individual occurrences of each word with the preceding and following stretches of text. For The National Corpus, we examined the broadsheets and the tabloids separately so that we could determine similarities and differences in their narratives.

Collocations were generated on the basis of the following criteria:

• Span of 5:5 – a window of five words to the left and five words to the right of the search word;
• Mutual Information (MI) score ≥ 6. MI is a statistical procedure widely employed in corpus studies to indicate how strong the association between two words is. It is calculated by considering their frequency of co-occurrence in relation to their frequencies when occurring independently in each corpus.
• Minimum frequency of collocation: 1% of the frequency in the study corpus.

1. Findings

Tables 2 and 3 list the keywords in the National (broadsheet and tabloids altogether) and Regional Corpora respectively, in relation to the general language corpus, grouped by theme.

Themes	Keywords
Sea	sea, ocean, coast, islands, marine, beaches, seas, oceans
Brexit	Brexit
Industry	fishing
Means of transportation	boats
Travelling	flights
Measures	lockdown, restrictions
Pandemic	covid, coronavirus, pandemic, virus
Politicians	Trump, Boris

Table 2:‘Keywords’ in the National Corpus in relation to the BNC2014_Baby, organised by theme

Themes	Keywords
Geographical references	Plymouth, Aberdeen, Norfolk, Cornwall, Hove
Sea	sea, marine
Means of transportation	vessel, vessels
Distance	offshore
Rescue service	coastguard, lifeboat, RNLI
Place	harbour
Division	chopped
Measures	lockdown, distancing
Pandemic	covid, coronavirus, pandemic

Table 3:‘Keywords’ in the Regional Corpus in relation to the BNC2014_Baby, organised by theme

Table 4 provides the keywords in the Regional Corpus in relation to the National Corpus, organised by theme. These keywords highlight themes which occurred with a relatively higher frequency in the Regional Corpus as compared with the National Corpus.

Theme	Keywords
Government	council
Rescue service	coastguard, lifeboat, RNLI
Place	pier
Geographical references	Plymouth, Aberdeen, Brighton, Suffolk, County, Norwich, Ipswich, Welsh, Southampton, Belfast, Hove, Durham

Table 4: Top 20 keywords in the Regional Corpus in relation to the National Corpus, organised by theme.

As discussed below, these keywords uncovered various ways in which the sea is represented in the UK press. Note the prominence of words referring to the Covid-19 pandemic and measures to control the spread of the virus (cf. the themes of Pandemic and Measures in Tables 2 and 3). This is not surprising since our study corpora are restricted to news texts published in the year of the pandemic (2020) while texts in the BNC2014-baby (that predates the Covid-19 pandemic) comprises a wide range of text genres and covers a longer period, from 2010 to 2017. This also explains the overuse of prominent politicians’ names (‘Boris’ and ‘Trump’) in the National Corpus (Table 2).

Finding 1: The narrative frequently represents the sea in terms of economic opportunities: resources, profit and job creation.

In both broadsheet and tabloid newspapers, prominence is given to negotiations around the fishing industry after Brexit (see Brexit and Industry in Table 2). Here the discourse revolves around the terms and conditions for European fleets to fish in UK waters (Extract 1), thus highlighting the relevance of the sea as an economic resource and the importance of protecting one’s economic rights in the current political climate. In addition to collocating with ‘Brexit’ in both subcorpora, the keyword ‘fishing’ collocates with types of sea transport (‘boat(s)’, ‘fleet(s)’, ‘vessel(s)’, and ‘trawler’), fishing gears (‘gear’ and ‘nets’) as well as words indicating some kind of restrictions (such as ‘quotas’, ‘illegal’, ‘rights’, ‘access’). There are also mentions of how the Brexit deal would affect fishing ‘villages’ and ‘communities’ in the UK and EU countries.

• But the EU has been clear that the price of access to its markets must be access for its fishing fleets to British waters (The Times, 30/01/2020).
• The row over fishing rights following Britain’s EU departure still threatens to collapse trade and security discussions after another week of wrangling ended in deadlock (The Sun, 22/11/2020).

References to economic resources are also seen through the collocations of ‘sea’ with ‘North’ across the three corpora (see Figures 4-6), which point towards mentions of oil production in the North Sea.

While national newspapers frequently mention the oil industry’s revenues, regional newspapers report on the discussions on the North Sea Transition Deal. The North Sea is also mentioned in relation to measures to curb the spread of Covid-19 on UK platforms and avoid job losses (Extract 4).

• By the mid-80s, the North Sea was providing 10% of the Treasury’s revenues, enabling Conservative government tax cuts, covering the costs of unemployment, and paying off a historic balance of payments problem (The Herald, 14/06/2020).
• THE safety of North Sea workers is at risk over oil industry plans to rip up a vital sector-wide agreement, a trade union has warned (Daily Record, 05/03/2020).

Interestingly, most mentions of the North Sea in regional newspapers (74%) come from the two newspapers based in Aberdeen (Aberdeen Press and Journal and Aberdeen Evening Express). References to renewable energy are prominent in the Regional Corpus only, as indicated by the keyword ‘offshore’ (see Table 3), which uncovered mentions of the generation of renewable energy through offshore wind farms (Extract 5). This demonstrates how the sea is considered as an important source of revenue (in particular via job creation) in communities that have traditionally depended on the sea for income generation. In contrast, national newspapers frequently mention mining of the deep sea for minerals (see the collocations of ‘sea’ with ‘deep’, Figures 4 and 5), especially in relation to campaigns to halt deep-sea mining given its serious environmental impacts (Extract 6).

• A predicted boom in North Sea offshore wind jobs was branded “a pipe dream” by union bosses after the Scottish Government admitted it only uses “estimates” of current employment figures in the sector. (Aberdeen Press and Journal, 20/02/.2020)

Sir David Attenborough has backed calls to halt deep-sea mining for minerals that are in high demand for use in items such as mobile phones. (The Daily Telegraph, 13/03/2020)

The Irish Sea is also frequently mentioned in the three corpora (see Figures 4-6). Here the discussion revolves around the negotiations between the UK and the EU in the context of Brexit as it entailed a regulatory border between Great Britain and Northern Ireland for the crossing of goods (Extract 7-8). It is interesting to note that 62% of the mentions of ‘Irish Sea’ in the Regional Corpus came from the Belfast Telegraph.

• Animal products arriving across the Irish Sea from Great Britain – including meat, milk, fish and eggs – will have to enter through a border control post where paperwork will be checked and a proportion of goods will be inspected. (The Daily Mail, 10/12/2020)
• But no matter how gently this was presented, an Irish Sea border has become a living reality, courtesy of Boris Johnson. (Belfast Telegraph, 16/12/2020)

Finding 2: The ‘marine environment’ (and not the ‘sea’ itself) is represented as a natural resource that must be preserved, protected, especially in view of sustaining the economic benefits from the sea.

The sea is represented as a natural resource to be preserved. However, this is seen through the collocations of the keyword ‘marine’ (see Tables 2 and 3) rather than through the word ‘sea’ itself. ‘Marine’ collocates with words such as ‘protected’, ‘conservation’, ‘ecosystems’ and ‘environment’ across the three corpora (Figures 7-9), uncovering mentions of initiatives and campaigns to protect marine life (Extract 9).

• Conservation charity Blue Marine Foundation called for one of the first pilot sites to be established in Wembury Bay, in the Plymouth Sound National Marine Park, to protect its varied marine life and habitats and to help connect people to the sea (Western Daily Press, 08/06/2020).

Concern about preservation of the sea ecosystem is also seen through the analysis of the keywords ‘boats’ or ‘vessel(s)’ (cf. Means of Transportation in Tables 2 and 3). The reporting revolves around campaigns against supertrawlers fishing in UK waters, especially in protected areas, as they have a negative impact on fishing ‘villages’ and ‘communities’. In the National Broadsheet Corpus, this narrative is also seen through the collocations of ‘fishing’ with ‘sustainable’ and ‘protected’, which unveiled references to sustainable fishing (Extract 10).

• Marine wildlife monitors say the vessels are destroying fish stocks, killing non-target species, harming sustainable fishing communities and destroying marine ecosystems (The Independent, 12/12/2020).

Finding 3: Newspapers frequently stressed the negative impacts of climate change on the sea. Climate change is a prominent theme in the discourse of both national and regional newspapers, as indicated by the association of ‘sea’ with ‘level(s)’ across the three corpora, and with ‘rising’, ‘rise’ and ‘ice’ in

the National Broadsheet Corpus (see Figures 4-6). The newspapers frequently mention the rising of sea levels due to higher global temperatures (Extract 11). The National Broadsheet Corpus specifically mentions declining of sea ice cover in the Artic Ocean due to climate change. This is evident through the collocations of the keyword ‘ocean’ with words such as ‘Arctic’, ‘temperatures’ and ‘warming’ (Extract 12).

• Sea level rises have accelerated in recent decades, threatening coastal areas and low-lying land by 2100 (The Express, 16/05/2020).
• Ocean temperatures in the area recently climbed to more than 5C above average, following a record breaking heatwave and the unusually early decline of last winter’s sea ice (The Guardian, 22/10/2020).

Finding 4: An ‘emotional’ lexicon can only be found in relation to aesthetic considerations.

The keyword ‘sea’ collocates with ‘view(s)’ across the three corpora (see Figures 4-6), uncovering descriptions of places with ‘panoramic’, ‘stunning’, ‘incredible’ or ‘superb’ sea views. This relates to value of a sea view in the hospitality sector (hotels, accommodations, restaurants, etc.) as well as in private properties (Extract 13).

• The property is spread over five floors and has key features such as a built in library, spectacular formal reception room and stunning sea views (The Argus, 12/09/2020).

Finding 5: A sense of place can be found in relation to the seashore/coastal locations.

A clear distinctive feature of regional newspapers relates to the prominence of place names (cf. Geographical References, Table 3). This is interesting because these draw attention to places close to the seashore. Through the analysis of the keyword ‘harbour’, we uncovered mentions of seaside towns with historic walled harbours where boats line up. Another feature specific of the Regional Corpus relates to Rescue services (Table 3) which refers to the services provided by local councils or charities in case of emergencies to ensure people’s safety on the beach or to rescue animals and objects from the sea (Extract 14).

• On average, the Redcar lifeboat is called out between 50 and 70 times a year, with the RNLI nationally being involved in nearly 10,000 emergencies in 2019. (The Northern Echo, 21/10/2020).

The set of keywords in Table 4 provides further evidence for the prominence of Geographical References and Rescue Services in the Regional Corpus. The occurrence of the keyword ‘council’ is not surprising given the composition of the corpus (local newspapers). There are several mentions of councils’ planning and services, including those to ensure safety in the seaside. The keyword ‘pier’ corroborates the trend indicated by the ‘harbour’ by highlighting people’s connection to the sea through sports and leisure activities in seaside towns and counties (Extract 15).

• Along with the pier’s existing aquarium and new rollercoaster, a £4m investment has added an indoor and outdoor adventure golf course as well as a children’s soft play area, covering more than 30,000 sq ft (East Anglian Daily Times, 01/02/2020).

6. Conclusions and recommendations

The dominant narrative in the British written press represents the sea as an economic resource and, at the same time, as a ‘marine environment’ to be protected and preserved. The sea is also recurrently represented as in needs of more regulations (especially in the context of Brexit and the migration crisis). We found some ‘emotional’ words in reference to ‘sea views’, but the dominant narrative is clearly one of utilitarianism and opportunism: in other words, the sea must be protected as it is useful, and not so much because we have any sense of belonging and connection to the sea. This fits with a weak conception of sustainable development that prioritises economic needs over environmental preservation (while trying to find a balance between these two necessities). The representation of the sea in British media demonstrates that there is an awareness of the benefits of the sea for livelihood and a need for the marine environment to be protected. But what is lacking is the conveyance of a real sense of place and belonging.

This has important consequences regarding ocean awareness. Indeed, having citizens worrying about the sea because they understand that the sea is economically important for them is certainly a good beginning, but this remains within the frame of a weak sustainability approach. Ocean sustainability (i.e. recognising the utter importance of the environmental and social dimensions of oceans) requires a stronger emotional connection with the sea. The narrative around the sea is too much utilitarian/opportunistic and not emotional enough. This contributes to a lack of sense of belonging and the valuing of oceans for their sole economic importance. In sum, our findings show that public policy stakeholders which want to further develop ocean awareness among the wider public need to contribute to the promotion of a narrative about the sea that is not just utilitarian (revenue, job creation) but also emotional.