"WHAT DO THE PAPERS
SAY?"
HOW PRESS REVIEWS LINK NATIONAL MEDIA ARENAS IN EUROPE Jessica erbe
Abstract
Numerous empirical studies have measured to what degree national mass media arenas in Europe converge in terms of news agendas, ways of framing, and reporting political claims from other countries or EU actors. This article suggests explaining such output-oriented findings with a theoretical model of mechanisms that link different mass media arenas. It is derived from national settings, where mass media arenas are also highly fragmented and yet sufficiently interlinked to be considered a public sphere. The most prominent mechanism is news agencies providing similar input to many arenas. The article draws attention to an under-investigated mechanisms of cooperation, mutual observation, and inter-referencing between different media outlets. Empirical evidence about press reviews in several European countries is presented here for the first time. The daily practice of radio stations, newspapers, or websites, to explicitly quote commentaries or news from other media is found to be a routine in many countries, and it often crosses national borders. The results suggest that the "opinion geography" of press reviews differs from the usual "news geography" in European media, frequently entailing a Europeanised perspective on whose views
are seen as relevant.
Jessica Erbe is guest researcher at the Wissenschaftszentrum Berlin fur Sozialforschung, e-mail: erbe@wz-berlin.de.
CM Oi
LD r-s
fN L/^
o o
fN CM
O >
A
LP
r->
Fragmentation of Public Spheres: A National and European Problem
The main research question of the article is motivated by a simple observation of pluralistic democracies guaranteeing freedom of the press: Many people read only one newspaper, or always have their radio tuned into the same channel, or they obtain news exclusively from TV. With so many different local, regional, and national newspapers providing political information at the same time, and countless other print and electronic mass media outlets doing the same simultaneously, how can one refer to only one public sphere? How can members of society obtain a minimum level of similar information on politics in order to define their opinion and preferences? How can they assume that other members of the same society using different channels have received at least a ground stock of similar input? We already know the most important reasons for the coherence that can be found despite the many channels from media studies; news agencies and shared news values among journalists represent the most prominent ones. It becomes apparent how much this multi-channel situation is in fact amplified when considering the European Union as a political entity with 25 internally fragmented national public spheres.
Political communication scholars in many countries have raised concern over phenomena of disintegrating public spheres (Jarren, Imhof and Blum 2000). Segmentation of publics is observed along the private interest lines of media consumers (thematic publics), political orientation (see Sunstein 2001 on the balkanization of the public sphere in the Internet), and other dividing lines. From a normative point of view, fragmentation of media publics is seen as a challenge because of the media's crucial role in respect of legitimacy of political power and societal cohesion in modern democratic systems. In pluralistic countries guaranteeing freedom of the press and information, multi-channel conditions for public communication are the standard situation. The emergence of electronic media has added further channels, although in the beginning of television broadcasting, the low number of channels available may have played a rather integrating than segmenting role, even more so as TV and radio reached parts of the population that did not consume print media. With deregulation of TV and subsequent multiplication of programmes, however, these electronic media have contributed to the fragmentation of the public spheres. The Internet has, at least for those who use it for political information and debate, exponentially increased fragmentation. One of the problems of political communication when multiple arenas are not interlinked is that groups may tend to wall themselves off from other viewpoints and that citizens will be exposed to only a selection of arguments rather than to "society's total argument pool" (Sunstein 2001, 74; Sunstein 2002). This imperfect information input may have an adverse effect on sensible political choices among citizens.
Since the mid-1990s awareness about the fragmentation of mass political communication in Europe has risen and nourished an intensive scholarly and political debate on the possible or impossible emergence of a European public sphere.1 In the light of more recent empirical findings on this subject, the approach followed in this article is to study the Europeanisation of public spheres rather than a European public sphere (see Schlesinger and Kevin 2000).
Mechanisms that Link Fragmented Arenas of Mass Communication
r-> ro-
under multi-channel conditions described above, democratic political entities need mechanisms that link the diverse political arenas and thus function as a "glue" for the fragmented public spheres. These bridges should be studied empirically at European level for assessing whether the 25 national public spheres in the EU are, as sometimes alleged, still hermetically separate from each other, or whether they are interlinked in a way that ideas, news, claims and opinions diffuse into many arenas simultaneously and the spheres are thereby integrated. The most important mechanisms are systematised in the model shown in Figure 1 and explained below, indicating existing research or desideratum. The model can be applied to any geographical entity or space, from the local to the global; and the more intense that these mechanisms are empirically present in a given space, the more one will be inclined to speak of a public sphere. This is a dynamic definition, which allows explaining such phenomena as transient public spheres, as for instance the global public sphere witnessed in the weeks before and after the war in Iraq in 2003. But the aim here is to provide a framework for exploring a sustainable Europeanisation of these mechanisms.
Figure 1: Model of Mechanisms that Link Mass Media Arenas
External input side		
M2 political actors address claims	f Political (	News ^
to several media arenas	actors J \	agencies J
M3 news agencies feed the same news to		
several media arenas		
Internal input side /		
M4: mutual cooperation, observation and *		
reference between media arenas / ,, ,. / Media	/ M4 (i-iv)	t / Media /
i formal cooperation between media / Arena ! /		/ Arena 2 /
ii mutual observation by media ^-'		' --'
iii inter-media reference as news sources		
iv press reviews	^-----.	
Reception side	{ Audiences ,	
M1 : some media users participate in more than	_	
one arena		
M1: Partially overlapping media audiences: Individuals often use more than one media arena for obtaining political information, either by combining different media types (e.g., TV and the Internet), or by using different outlets of the same media type (e.g., reading two or more newspapers). In this way, subscribers to different newspapers as well as people who do not read at all, and who would receive very diverse information and opinions, may "come together" and form a common audience of prime time national TV news. While such linking of arenas by the audiences themselves is an everyday reality at national level, it is strictly limited, at European level, to an information elite, people with a migration background, or to users of transnational media exceptions such as Arte. This lack of overlapping au-
diences at European level actually gave rise to the pessimism about the emergence of a European public sphere in the political and academic debate in the 1990's.
M2: Same news input into different media arenas by speakers: Collective actors such as parties, governments, civil society and interest groups actively seek to introduce their claims into the media in order to gain support. For reaching many media arenas simultaneously, they give either indirect input through press agencies or direct input through press conferences and press information distributed to a large number of news rooms, or through targeted telephone calls to journalists. Regarding the Europeanisation of public spheres, it is necessary to explore to what extent this national practice is also carried across national borders, a research question addressed in the project Europub.com (Koopmans and Statham 2002).
M3: Same news input into different media arenas via news agencies: Already at national level, news services make up a considerable part of the daily printed material. In particular, the regional newspapers (at least in the case of Germany) feed a large part of their non-local or non-regional news pages with agency material. As agency texts have become more and more ready-to-use, readers of hundreds of regional and local papers throughout Germany will in the end have received, without being aware of it, a considerable common set of news stories, sometimes even in the same wording. This phenomena, although elsewhere discussed as problematic at national level in terms of informational freedom, and at international level in terms of hegemony (see Biltereyst 1995), can be considered a very important bridge between the many different publics.
M4: Inter-media cooperation, observation, and reference: There are four types of mechanisms practiced by media.
1.	Formal cooperation between mass media: Highly competitive media markets normally exclude formal editorial cooperation between outlets of the same media type at national level. This is much less the case for cross-border cooperation, since markets for quality broadsheets are still confined to national borders.2 Inter-media cooperation can lead to the transmission of identical content to otherwise separate audiences, and thus functions as an interlinking mechanism. Less direct forms of inter-media cooperation, for instance the exchange of journalists, may also do so, but to a lesser degree.
2.	Inter-media observation: Invisible to their audience, mutual observation of media producers is taking place every day (Pfetsch, Eilders and Neidhardt 2004). TV editorial departments scan the press while print media journalists follow TV and radio reporting. Also competing outlets of the same media type observe each other and are influenced by the others' behaviour, selection of issues and evaluation and framing of issues, a practice also analysed in agenda setting research. Further research is needed on the extent to which this strong mechanism is also practiced across national borders.
3.	Inter-media reference as news sources: In some cases this observation becomes visible to the audience by explicit reference to another media (for instance Les Echos referring to an exclusive interview with EU Commissioner Bolkestein in Handelsblatt, or to results of investigative action of a TV crew). Here, the core information is diffused into several arenas; other outlets are quoted as information sources just as news agencies are quoted. Research on the Europeanisation of this mechanism was initiated by Scherer and Vesper (2004); future research should cover more countries and focus on the ways such references are produced, as in many
cases, so-called pre-information on exclusive interviews is systematically distributed to other media through press agencies.
4. Press reviews - opinion-related inter-media references: Media are acknowledged as political actors of their own stance in many media studies (Page 1996); they try to influence politics through opinion formation, and in particular through their commentaries and on op-ed pages (Pfetsch, Eilders and Neidhardt 2004). In the journalistic genre of press reviews, media explicitly quote each other's opinions or main preoccupations. This practice is an intentional linking of different public arenas; a miniature public sphere is created within the same page of a newspaper or within ten minutes of a radio programme. This practice still lacks research both on the national and transnational level (except for Le Bart 2004 and Scherer and Vesper 2004).
For a comprehensive assessment of the extent to which media arenas in Europe are inter-linked, information on all of these four mechanisms should be collected. The fourth among them is still under-explored, and in particular in its variant of press reviews, that is how media in Europe refer to each other as participants in a public discourse. This article aims at raising awareness for this mechanism and provides a first cross-national empirical account of it.
"What the Papers Say" - Typology and Production Process of Press Reviews
In this article, the term press review is defined as the practice of mass media to present the opinion or main preoccupation of other mass media by either a literal quote or a commented summary. This does not include the reference of mass media to other media outlets as a source of information (e.g., La Stampa reporting that according to the New York Times' reports, the US administration plans to...). It also excludes the manifold professional press clipping services that provide public authorities, private companies, interest groups and non-governmental organisations with targeted press clippings. Only regular practices are considered, i.e., when references to other media are made either with regular periodicity (for instance each issue of a daily or weekly paper, or each morning in a radio programme), or in a consistent format (not necessarily every day, but in a uniform, recognisable format, for instance, on the opinion page of a broadsheet whenever the editors judge an issue appropriate for presenting outside views).
Empirical research on press reviews and particularly on their European aspect has only been initiated very recently and has thus far been limited to a qualitative analysis of one specific French radio case, namely Alex Taylor's former "revue de presse européenne" on France Inter (Le Bart 2004) and a quantitative analysis of two German newspaper cases (Scherer and Vesper 2004). Because of this lack of basic information this section offers a typology for better understanding the phenomena.
Three main types of opinion-related press reviews can be distinguished. The individual voices review reproduces a literal extract of one or several editorials or commentaries. Each quote is preceded by some meta information, such as the source's location, editorial line or type (tabloid, national reference paper, etc.), the name of the commentator3 if signed, and a short indication of the text's tenor. The commented joint review cites the reasoning of the source in less detail; sometimes
only the headline or the framing of a political event is literally quoted. Instead, positions of several sources are summarised on the same subject, explained with background information, or even commented upon. This type involves own editorial work, which at times is styled so as to tell a story. Compared to the other type, its production therefore requires more resources; on the other hand, its output can be more dialogical and thus more entertaining. The disadvantage is that readers will see other media's opinions not fully argued and only through the filter of their own media's journalist's opinions. With the Internet developing into an additional forum for political discourse, a third type has arisen: the linked index review. It does not present detailed contents of other media's opinions but - using deep hyperlinks to the text - only where these are to be found online (e.g., the EU affairs press review on Europa-digital.de). Its added value consists in the selection of relevant sources among the vast number available, and perhaps in the indication of the subject and the stance of the quoted opinion. In practice, some newspapers and especially radio stations publish a written version of the press reviews they printed or broadcasted on their website.
In fact, how do press review editors obtain the original texts? They can combine several sources; besides the traditional reading and copying of the actual paper edition, they consult the websites of leading titles and draw on ready-for-use services from news agencies. German editors receive translated excerpts of editorials of other media through the dpa service, and LW from both dpa and AFP. In special cases they call their correspondents abroad. A practice worth exploring is that many newspapers systematically transmit by facsimile their of opinion articles not only to the news agency but also in full length to many other media outlets, in particular to radio stations. This witnesses a real interest in the diffusion of their opinions, be it for competitive purposes (being quoted is an indicator of prestige) or for reaching a wider audience with their political views. A combination of both motives seems most likely. If this practice was to be carried across national borders, if there was a regular exchange of editorials say between Liberation and Pravo in Prague, Dagens Nyheter in Stockholm, and Delo in Ljubljana, this would be a truly Europeanised media dialogue.
Functions of Press Reviews in the National and European Settings
As explained above, press reviews constitute intentional linkages between media arenas, and their primary role is to inform media users of what is being thought and debated elsewhere, just as news do. If the four-million-reader tabloid Bild launches a campaign against a politician, other German media can hardly ignore this. Since newspaper commentaries are seen as an indicator of (part of the) public opinion, it is important to have an overview of their respective tenor for anyone interested in politics. Professionals in the field obtain this information via press clipping services, the general public via press reviews. At the same time, their role is to contribute to the opinion formation among their users. By hearing or reading a whole range of possible opinions and arguments, media users obtain a richer knowledge base for their opinion formation than if they were to receive only the views of a single media outlet. They might develop a better understanding for the reasons that make a political camp take a certain position, even if they continue to disagree with this position. From a normative point of view, press reviews thus contribute to "reasoned" deliberation. A different approach would be to consider their
"communicative utility," a concept leaning on the uses and gratification theory (McDonald 1990). In this sense, individuals would use press reviews in anticipation to inter-personal communication, say with colleagues, where the additional arguments help increase prestige.
In any case, thanks to the fact that they visualise the other participants in a public sphere, press reviews have an interesting potential to contribute to the Europeanisation of audience's consciousness about the public sphere. Research on the emergence of nation states has shown how much political communication has contributed to such processes (Anderson 1992). Media discourse - which usually claims to simply acknowledge existing territorial boundaries - in fact often contributes to reinforcing these very boundaries, as scholars argue both for the regional level (Ringoot and Utard 2004) and for the national and European level (Le Bart 2004). In applying this assumption to the topic of this article, one may argue that inter-media references repeated over time have an impact on their audience's conception of the public sphere. If on all policy matters, from schooling to climate change, EU chemical stuffs directive to military intervention in Afghanistan, German audiences hear or read exclusively German voices, their intuitive knowledge of the public sphere, and thus of who else is involved in a communicative space with them, risks remaining exclusively national. If, on the contrary, they are used to hearing in their radio or reading how Danish, Slovenian, and Italian media frame certain issues, they may develop a different perception of who else is concerned by the political process. Obviously, awareness about the other participants in the public sphere must also be raised by other media practices, above all by political actors from other European countries staging claims on common policies in the media (Koopmans and Erbe 2004).
Research Design
The empirical findings presented in this article draw on a larger set of data collected for my PhD project, which focuses on the German case. Data on other countries were compiled mainly for comparison, and are thus less detailed. The study consists of two empirical parts: The first relies on the method of own observation of mass media outlets in many countries, combining the use of newspaper archives, own listening and viewing of programmes, studying radio and TV programme schedules in newspapers and specialised magazines or on websites of the respective stations. To carry out such research systematically for the multitude of mass media outlets in each country would require extensive research resources and a multilingual team, exceeding by far the scope of this project.4 The observations were made from 2003 to 2004, but as reorganisation of newspaper pages and radio schedules makes them a 'moving target', occurrences found might reflect only parts of this period.
The second part of the study is focused on content analysis of press reviews of selected cases from four different countries or regions: Spain's prestigious newspaper El Pais (EP), Luxembourg's most widely read quality paper Luxemburger Wort (LW), an aggregate of German quality newspapers (DE), and as a European transnational case, the Brussels based weekly European Voice (EV), whose readership is limited to EU officials and lobbyists in Brussels and which is seen as fairly influential in European affairs. The German case is an aggregate of the two leading national quality papers Frankfurter Allgemeine Zeitung (FAZ ) and Süddeutsche Zeitung
(SZ) as well as of two Berlin based quality papers, tageszeitung (TAZ) and the regional Berliner Zeitung (BlZ). The average of these papers is obtained by weighting according to the number of sold copies of each title, as provided by Schütz (2001). Besides the two large countries Germany and Spain and the genuinely European case, the Grand Duchy was chosen as a small country, because of the interest of the multilingual audience it serves and because one may assume that it is being influenced by both French and German political and media cultures. For all seven newspapers, a full population content analysis of all press reviews published in the reference period 2000-2002 was carried out, the unit of analysis being the individual media opinion quoted in these reviews. The number of opinions quoted per day varies considerably. In the single-voice type reviews, it ranges from only one in EP to seven quotes on average in FAZ, LW quoting 3.5 on average. In the commented joint review of EV, a quote was counted whenever a media outlet was explicitly referred to; if the same source was cited on a different subject matter, it was counted again (such double or triple mentions account for less than 4% of all quotes). For each opinion quoted, the date, name of the quoting media, name of the quoted media (source), and country where this source is published, were recorded using a codebook of 1000 media titles and entered into a database. For the purpose of this article, the values were then aggregated by country of origin of the quoted sources, so as to obtain a quantitative measure of the attention that a specific press review gives to each foreign country. Due to limited access to archives, the entire 3-year reference period is covered only for the four German dailies; the last three months are missing for LW, the last 12 for EP and the first 12 for EV. Depending on the period of reference, the original newspaper content was retrieved either from online archives and CD-ROMs, from paper copies or supplies by the editor (LW). As press reviews appear in regular sections, there was no need for using full text search terms. The total number of quotes (N) obtained is 19327 (LW: 2576; 4 DE papers: 15403; EP: 606, EV: 742). All remaining information was gained from interviews with editors and radio programme managers.
Hypotheses and Indicators
The first step, which can only be introduced in this article, would be to compile an inventory for each country listing major national and regional dailies as well as radio stations, and to determine whether or not they practice a regular look into what other papers say. The next step is to examine to what extent those media in Europe that offer press reviews to their audiences also take opinions of foreign media into account. This can be measured through a simple indicator, the degree of internationalisation, which corresponds to the ratio of quotes from foreign newspapers in total quotes of a given period. The degree is 0 when only domestic voices are mirrored, and 1 when all quoted media are from foreign countries. Another indicator to be measured is the degree of Europeanisation. It is defined as Europe's share in the total number of foreign quotes compared to other World regions and can range from 0 to 1. Given the ambiguity of the term "European," an in-depth analysis should rely on a set of several degrees of Europeanisation, but for the purpose of this article, only the simplified measure of the continent is calculated.5 As there is only a certain amount of space available for the daily press review, selecting opinions is a zero-sum game: each privileging of own national sources will be to the detriment of foreign views, and focusing mainly on European opin-
ions means ignoring other continents' views. The last measurement is a combination of two indicators describing the geographic focus within Europe. One measures the width of coverage, i.e., the number of different European countries covered. In the wide definition of Europe used here, it ranges from 1-35 out of 35 countries. The other indicator returns the concentration, defined as the cumulated share of the four European countries whose media opinions are quoted most often, in terms of total European quotes. If a press review focuses on only three or four main European players or neighbours, it has a high concentration. For the theoretical argument of awareness raising (see above), a low concentration of a press review would allow its audience to better understand who else is concerned about a political issue. On the other hand, such understanding is probably based on recognition, i.e., a cognitive base on the side of the reader or listener. While the names of prestige papers from large countries such as Times, Monde, FAZ, Corriere or El Pais are widely known, it is almost inevitable that with an increasing width of coverage, the chances for an individual newspaper to be referred to repeatedly and thus to become a familiar name, are diminishing. Finally, these two indicators of the geographical configuration within Europe need to be qualified by non-numerical information on which countries are primarily covered.
These measures can be considered as indicators of the prominence that countries or regions hold, manifest in the attention that their media receive in foreign media. They describe the communicative space that press reviews open for their readers or listeners. What would one expect this space to look like? There are no theories or empirical analyses as yet on the configuration of opinion-related media content.6 Hypotheses must therefore lean on theories on news-related media content, namely the news flow and news geography theories (Hagen et al. 1998; Sreberny-Mohammadi and Stevenson 1999; Wu 1998). According to these, news values - the criteria that guide journalists in the selection of news - are not only related to what happened but also where it happened, implying there are characteristics of countries that make similar events occurring in different countries more or less likely to be reported. Such country characteristics are above all status (political, economic, military, etc.) and closeness (political, economic, geographic, cultural, etc.). The United States, which "emerged as a unique news superpower" according to Stevenson's preliminary account of the major post-Cold War comparative study of news geography in more than 40 countries (Stevenson 1998, 117) are found to play a dominant role in most studies. One can therefore expect US media to receive a great deal of attention from press reviews in any country. With regard to the second news value, closeness, one would expect newspapers from European countries to receive a lot of attention as well. Here, some variance can be expected between the press reviews under scrutiny, as their respective countries differ in terms of neighbourhood and cultural (including linguistic and historic) ties. Opinions from a few other big global players might as well attract interest, for instance from the former superpower Russia, from Japan as a main trade partner, from China and India as large, fast growing economies and holders of nuclear weapons and major populations. Views from media of other countries would be expected to be consulted on an event-related (mainly crises) rather than on a regular basis. Another factor that might influence the chances of an op-ed to be mirrored in another country is language: Since press reviews are published in the language of their respective audience, they need to be translated,7 involving fi-
nancial costs and a delay in time. Thus the cost of translation might be an obstacle to the cross border diffusion of opinions, a structural obstacle that can best be overcome by collective services such as news agencies. With respect to the degree of internationalisation, the size of the country where the press review is edited might play a role, but also its relative dependence on foreign trade; the hypotheses being that small countries have a relatively greater interest in media views from other countries, and that countries with a high share of exportations in GDP also give foreign media voices more space in their press reviews.
Findings: Press Reviews - A Highly Europeanised Media Discourse
In the case of outstanding events, the news media we use tend to not only report and comment the event itself, but also the way it is framed by other media -the election of Pope Benedict XVI being such an event. Besides such event-related instances, many media outlets throughout the world also practice regular press reviews. This section aims at starting off a systematic inventory of this practice in Europe.
Regular press reviews are a journalistic genre known in most European countries. They are most common in radio programmes with a high information share. The routine look into what the papers say is made in the highest-audience hours in the morning, thus informing listeners before they leave the house or when they drive to work. In France, three of the four radio stations with the highest audience for morning programmes, the public France Inter and France Info, and the private Europei broadcast detailed daily "revue de presse", and so do the public France Culture and the foreign service station Radio France Internationale (RFI). The German public broadcaster Deutschlandfunk/DeutschlandRadio, which reaches an audience of millions, transmits several press reviews ("Presseschau") per day including international voices, and in each of the nine regional public broadcasters there is at least one programme with press reviews (Erbe 2005). In Spain, the two most important radio stations La Cope with "resumen de prensa" and La Ser with "revista de prensa" offer a national and an international one on every workday. Other examples of regular inter-media references in the radio are the BBC, the Belgian BRF and Swiss DRS1 and DRS3.
In print media, the spread of this genre is less systematic. It is most widely practiced in Germany, where it has a long tradition, but also more generally in the German-speaking press culture, including media in Austria (Die Presse), Luxembourg and Belgium (Grenz-Echo, and francophone Libre Belgique). It is a common feature in the Spanish press as well, e.g., in El Pais, La Vanguardia, and Cinco Dias. Le Monde reviews French press and TV in "dans la presse frangaise" and on the same page foreign media in "vu par la presse..." edited by Le Courier international8; similar cases were found in Le Figaro and La Tribune. In the United Kingdom, the most prominent example is the Guardian's section "The Editor," a whole page devoted to "the best of the world's press digested for you"; The Express is another example. In The Netherlands, press reviews are only published sporadically in case of a special issue.
Television is not a suitable medium for press reviews with their focus on words, and there are only few exceptional cases to be found. In earlier years German public TV broadcasted the editorials of major newspapers (in order to let, as sharp tongues said, East Germans participate in Western media discourse), and in the
reference period Euronews also read out headlines of major European newspapers, showing the original text - also in Cyrillic alphabet when quoting Kommersant from Moscow. German foreign-service broadcaster Deutsche Welle TV also used to include press reviews - an "anachronism" according to a then DW-journalist.
On the contrary, the Internet with its hyperlink technique and less space restrictions provides excellent conditions for the genre. Besides such purely virtual press review editors as the renowned Perlentaucher.de, which focuses on cultural and media pages in newspapers, the main providers are the online departments of the major print media and radio outlets worldwide. La Cope, RTB, RFI, and DeutschlandRadio, for instance, present the entire text of their reviews online, in some cases hyper-linked to the home page of the quoted source. The two latter offer these texts as free-of-charge newsletters. A fully edited review with the enthusiast title "les hyper-nouvelles de la planète" was provided during two years on the French daily Libération's website. National borders are crossed with a simple mouse click.9
Overall, press reviews prove to be a regular feature in political communication in numerous mass media in Europe. Their practice depends, however, on the specific media outlet, and cases without press reviews outnumber those with. In the radio, it is those programmes used by people interested in political information that usually broadcast an overview on what the papers say. Technological change and habits of media use make the genre disappear from one media type, TV, and appear in another one, the Internet. After these preliminary observations on the spread of regular opinion-related press reviews in Europe, it is worth examining the communicative space that they build for their audiences, i.e., the geographical set up.
The selected cases of press reviews show different degrees of internationalisation. As the title of the Luxemburger Wort's section "Auslandspresse" suggests, only foreign media opinions are quoted and the few other media voices in Luxembourg are omitted. The outward orientation represented in this degree of 1.00 confirms the hypothesis that small countries have great interest in media views from other countries. The example of El Pais as a leading paper of a large country, however, refutes this hypothesis, as its "Revista de prensa" also reaches a degree of 1.00 with only two Spanish papers cited. For Craig Winneker's section "What the papers say" in the European Voice, the degree of internationalisation depends on how one defines its constituency. As EV is hardly read outside Brussels and has its editorial staff in Brussels, one might consider all quotes from non-Belgian papers as foreign. The UK-based ownership (The Economist Group) and predominantly Anglo-Saxon crew and journalistic style would offer plausible reasons to considering only British papers as fellow national sources. In both cases the degree would be 0.9. However the weekly's claim to be European and its multinational EU readership as well as its self-image, may as much justify treating only transnational papers as belonging to the EV's own political realm10 (degree of 1.00), as considering all media in Europe as from EV's 'own country' (degree of 0.11, i.e. only 11% non-European opinions). In the weighted average of the four German cases, the degree is 0.49, thus half of the press review is devoted to non-German media. Among the four, FAZ gives least room to foreign views (0.38), Süddeutsche and Berliner Zeitung half (0.53 and 0.51, respectively), and TAZ most (0.83). Considering that press reviews furthermore play an integrative role in their own fragmented national pub-
lic spheres, a threshold of one third of the total space devoted to foreign views is already a considerable degree of internationalisation. All seven newspapers examined here exceed this threshold. However this does not necessarily hold true for press reviews in Europe in general; some regional papers in Germany for instance focus more on relaying their regional audience to the federal media debate than on an international debate.
So, how is this foreign share in press reviews made up? If one groups the quotes by world regions where the cited sources are published, it becomes clear that there is a highly Europeanised media discourse led in press reviews (see Table 1). The highest degree of Europeanisation - in its wide definition of the continent excluding Russia and CIS - is held by Luxemburger Wort (0.94), followed by the four German papers (0.91, ranging from 0.89 in TAZ to 0.93 in FAZ) and the Brussels paper (0.88). Compared to these, El Pais is least Europeanised (0.51), but still devotes more than half of its press review to other European countries. When using a narrower definition of Europeanisation and counting only the 25 EU members (during the reference period the 10 countries to join in 2004 appeared already as potential or future members in media discourse and should therefore be included), the results are still valid: LW: 0.84, DE: 0.88, EV: 0.79, and EP: 0.50. The difference stems mainly from Swiss and Norwegian media, which attract fairly much attention.
Table 1: Degree of Europeanisation of Press Reviews, 2000-2002
	Quotes by world region (in per cent of total quotes of media opinions from foreign countries)				Total = 100%
Newspaper	European countries (except own country)	Russia	USA	Rest of the World	
Luxemburger Wort	94.2	5.1	0.2	0.6	2,576
German newspapers	91.3	7.0	0.5	1.3	15,403
El Pais	50.6	0.3	24.0	25.1	611
European Voice	88.5	0.7	9.7	1.1	746
Note: "German newspapers" include (weighted average of) Frankfurter Allgemeine Zeitung, Süddeutsche Zeitung, tageszeitung (TAZ), and Berliner Zeitung.
The most striking result of this empirical analysis is the absolutely marginal share of US media in the four German and in Luxembourg's press reviews, where it is less than one per cent (0.2% in LW and 0.5% on average in DE). This lack of interest in the media debate in the worlds' politically and economically most powerful country is somewhat surprising. Naturally, in order to be selected by a press review editor, the foreign opinions must deal with questions of interest to the respective audience, but one would assume sufficient opinion articles in US-American media dealing with global or diplomatic problems, which are of interest to Germany's and Luxembourg's audiences. After all, that 9-11 and the war in Afghanistan occurred in the reference period. The fact that El Pais and European Voice found indeed more US opinion articles worth quoting (24.0% and 9.7%, respectively) suggests that the reason for this disregard lies in the selection processes or
criteria in Germany and Luxembourg rather than in the offer side. Whether it is a decision made alone by the actual press review editors or partly by the news agency providing pre-selected material, can only be determined in interviews to be held in the future. Longitudinal analysis of German press reviews shows that this relative indifference to US media comments has developed over time, as these used to play a major role in the mid-1960s (Erbe 2005).
The German papers and Luxemburger Wort find views from Russia - the former super power and current veto holder in the UN Security Council - much more relevant (7.0% and 5.1%) than the Spanish and Brussels papers do (less than 1% each). Newspapers with a mainly East German audience reserve an even higher share of their press reviews to Russian media than those with a West German audience (Erbe 2005). A hypothesis for explaining this phenomenon is that former political adherence to different power blocks has influenced the intuitive political geography of media and audiences in the two parts of Germany. This "geography inside peoples heads", as Crang (1998) calls it, has apparently also led to a different set up of the communicative space in press reviews.
Media from other parts of the World, icluding other important economies such as Japan and Canada, the permanent Security Council members (China), and nuclear weapon holders (India, Pakistan), current areas of conflict (Israel, Middle East), and hundreds of other nations "compete" in the selection process for the remaining portion of foreign voices.11 But this share is negligible in all seven newspapers except in El Pais, where it constitutes one quarter of the foreign quotes. EP's attention for the rest of the World is largely due to the frequent quotes from Latin-America, which can be explained by the factor closeness, as the former colonial power Spain still has privileged relations to these countries.
Despite their high degree of Europeanisation, the press reviews do not link media arenas in Europe in a balanced way. In general, peripheral partners among the EU members such as Ireland, Portugal, and Greece are hardly ever listened to; and except Czech, Polish, and Hungarian views, the new members receive only marginal attention. The media discourses mirrored best into other countries are obviously those of the large and long-time EU members United Kingdom, France, Germany, Italy, and Spain; but there are specific cross-border focuses as well. The width of coverage is highest in the European Voice, with four fifths of the 35 countries covered at least once, followed by the German papers with three fifths on average and LW still covering more half of them. EP quotes views from only ten out of the 35 countries, partly because with one citation per day, one can cover fewer countries than big press reviews, partly because of the paper's transatlantic or even global focus. In the "Revista de prensa," the EU seems composed of five to seven countries and there is no indication that Eastern enlargement is ahead. EP also reaches the highest concentration within European quotes with 96% held by the first four countries, against 71% in the German average, 60% in LW and only 51% in EV. Instead, European Voice links to the media arenas of the medium sized countries as well, omitting only the Grand Duchy, Malta and Slovenia among the present EU members; thus it truly lives up to its name. The four German papers converge in their preferences for specific newspapers, La Repubblica, Le Monde, and Guardian are the top three opinion makers (see Pfetsch, Eilders and Neidhardt 2004 on the concept) in each of their press reviews. The catholic LW has similar preferences, but gives priority to the conservative Le Figaro, cites NZZ more frequently
than others, and is the only one to make the catholic La Croix an opinion leader. El Pais mirrors above all the views of Anglo-Saxon media, led by the New York Times, Economist and Financial Times, while the European Voice leans most on the commentaries of Le Monde, EP, and De Standaard from Brussels.
Conclusions
The European Union will have to work with the multilingual and fragmented public spheres that it is currently based on. But as fragmentation is also the reality of most national public spheres, this setting should not be regarded as an insurmountable obstacle to the Europeanisation of public spheres. This article proposes a model of the mechanisms that usually integrate fragmented national spheres, and indicates how they can be applied to the European level. Besides well-known mechanisms like news agencies, which feed the same information into many media arenas, the model includes several practices such as formal cooperation between media outlets, constant mutual observation by journalists, and opinion-related press reviews. In regard of the latter it is argued that this journalistic practice of explicitly quoting the commentaries of other newspapers helps linking otherwise separate audiences. As the lack of cross-national media discourse has often been deplored in the debate on European public spheres, it seems promising to explore the integrative potential of press reviews that would take into account not only domestic, but also other European media views. The potential is of course to be seen in relation to the limited status that this practice has in total political communication in mass media - it can merely constitute a small piece in the overall picture sketched in the model.
As there is hardly any academic knowledge on which to draw in this field thus far, the article first provides a definition and typology of this genre and explains the journalistic production process. An interesting finding is that the daily look into what the papers say is far from being random - on the contrary, some news agencies are busy everyday to collect, translate and systematically distribute the editorials of domestic and foreign newspapers to the editors of other media.
A first empirical overview suggests that in most European countries radio programmes or daily newspapers actually either make use of this material, or collect themselves other media's opinions in order to inform their audiences. Every morning, millions of radio listeners in France, Spain, Germany, Switzerland, etc., hear not only the news, but also how the leading newspapers react to them. And, as the seven case studies of regular press reviews from Spain, Brussels, Luxembourg and Germany, show, this practice includes a considerable part of cross-border reactions. Their editors prove to be very international and highly Europeanised in their selection. There are noticeable differences in geographical perspectives that the press reviews offer their readers. The Spanish one takes a rather Westernised than Euro-peanised look with a strong transatlantic focus, Luxembourg's and Germany's are almost entirely Europeanised, to the point of blinding out US media opinions, and the transnational case of European Voice appears as a 'model' European press review, including also some of the otherwise ignored media voices. A striking result is that news geography, for which there is a well-established research tradition, is not identical at all with opinion geography studied here for the first time. While news stories in European media focus on their region, and on the US as well as on crisis-
hit world regions, the framing of the news is reserved mainly to fellow European media (except for El Pais). To give an example, German press reviews prefer to 'observe' the US presidential elections through the eyes of Le Monde and Guardian rather than directly through US media. As a summary, press reviews are thus found to be a forum where Europeanised media discourse is already led every day.
As is argued in this article, one of the important roles of press reviews in the EU context is that they are a 'visible' link between media arenas in different countries. By citing foreign editorials on issues debated also at home, they visualise the media agenda of other countries. One criterion for Europeanised public spheres as formulated in many other studies is that the same issues be debated at the same time in similar ways throughout Europe. Here, the emphasis is on the actual visibility of such a potential similarity. By learning that their neighbours discuss the same problems, domestic audiences may eventually develop a better awareness of who else is involved in a joint communicative process. Such awareness is certainly necessary for the emergence of a public sphere.
Further research is required for exploring systematically what media in Europe are involved in this European opinion formation process, are what are the main factors enhancing or braking the process.
Acknowledgement:
The author wishes to thank the participants of the conference "The European Public Sphere" held in Berlin in November 2003, as well as the members of the Social Science Research Centre's Working Group "Political Communication and Mobilisation" for valuable comments on first drafts of this text.
Notes:
1.	See Javnost—The Public, issue 1 [2001] on "The European Public Sphere: Dreams and Realities."
2.	Towards the end of the 20th century, such cooperation has increased, either by exchanging articles, building multilateral alliances with other media (e.g., European Dailies Alliance and International News Alliance), or in more explicit forms such as regular columns of foreign media representatives or guest commentaries of foreign journalists. The most substantial form of cooperation consists in full-page reprints of editorial contents from partner media, usually as a supplement to a national daily. Le Monde diplomatique for instance is entirely translated into 17 languages and distributed in a large number of countries. Other examples are the 12-page supplements "FAZ weekly - Germany in English" in International Herald Tribune, and the NYT digest distributed in English as a supplement to major papers such Le Monde, Süddeutsche, etc. An example of inter-media cooperation among large European public service broadcasters for abroad is the project Radio E.
3.	A cultural difference worth noting is that German speaking media usually quote editorials and commentaries as the opinion of the respective papers, although copyright law requires the author's name. On the contrary, French media and also El Pais attribute comments to the journalist whenever the text was signed. This might eventually reflect a stronger role of the individual in the discourse tradition in France, where so-called public intellectuals are more institutionalised than in Germany.
4.	In addition, radio press reviews are only sometimes organised as autonomous contributions that can be found under such label in programme schedules. A trend away from small-portion programming can be detected towards moderated, several hour-programmes where news and elements such as press reviews are integrated in a fluid, flexible manner. In such cases, reliable information can only be collected through own listening and/or consulting programme directors, done thus far only for the German case. Newspapers do not necessarily publish press reviews
on a daily basis; for exploring whether a specific newspaper practices this or not, it is necessary to scan several issues, for instance over two weeks.
5.	Included are 25 EU members, as well as Bulgaria, Romania, Turkey, Norway, Switzerland, Albania, Bosnia, Croatia, F.Y.R. Macedonia, and Serbia/Montenegro. As quotes from the own country do not count, this total of 35 is reduced by one (except for EV, which can effectively quote media from 35 countries). Russia is not included here because of its former super power status, although it falls under the geographical definition of Europe. The other members of the Commonwealth of Independent States (CIS) are also excluded. From the former Soviet republics, only the three Baltic countries are included in the category 'European countries'.
6.	One of the main criteria when selecting commentaries for a press review is certainly that they take a clear stance on a current issue seen as relevant by the editor and that how they argue can be appropriately presented in a small excerpt. However, other criteria must also play a role, otherwise there would not be as high a concentration on certain countries and media outlets, repeated both over time and across reviews. The prestige of a source - as well as the perceived relevance of its country of origin - clearly influence the selection.
7.	The only exception is the Luxemburger Wort, which publishes articles in two to three languages anyway, and which can afford to reprint both German and French opinions in original language. Only for the linked index type reviews, translation is not mandatory: the editors will scan as many language areas as they are proficient in, and the users will follow links to original texts according to their faculties.
8.	Meanwhile, this section is moved from .horizons-kiosque" to the media section, the international part seems to be discontinued.
9.	To mention just two examples from non European media: The Lebanese Daily Star online provides both an Arab and an Israeli press review, and in a formal cooperation, New York Times and the influential news magazine Der Spiegel publish each other's editorials (both in English) on their respective websites.
10.	The transnational European web sites EUobserver.com and The Sprout were cited. In the wider definition, one might also include International Herald Tribune and Wall Street Journal Europe (otherwise counted as US media), but even then transnational media constitute less than 3% of all quotes (0.03).
11.	Freedom of the press is subject to more or less severe restrictions in many countries - a fact that might make the views published there less credible for press reviews in other countries.
References:
Anderson, Benedict. 1992. Imagined Communities. Reflections on the Origin and Spread of
Nationalism. London: Verso. Biltereyst, Daniel. 1995. Qualitative Audience Research and Transnational Media Effects. A New
Paradigm? European Journal of Communication 10, 2, 245-70. Crang, Mike. 1998. Cultural Geography. London: Routledge.
Erbe, Jessica. 2005 in press. Integration der politischen Öffentlichkeit in Europa durch Vernetzung. Der Fall der grenzüberschreitenden Presseschauen in Deutschland In W.R. Langenbucher et al. (eds.), Medialer Wandel und Europäische Öffentlichkeit. Wiesbaden: Verlag für Sozialwissenschaften (VS). Hagen, Lutz M., Harald Berens, Reimar Zeh and Daniela Leidner. 1998. Ländermerkmale als Nachrichtenfaktoren: Der Nachrichtenwert von Ländern und seine Determinanten in den Auslandsnachrichten von Zeitungen und Fernsehen aus 28 Ländern. In C. Holtz-Bacha et al. (eds.), Wie die Medien die Welt erschaffen und wie die Menschen darin leben, 59-82. Opladen: Westdeutscher Verlag. Jarren, Otfried, Kurt Imhof and Roger Blum, eds. 2000. Zerfall der Öffentlichkeit? Wiesbaden: Westdeutscher Verlag.
Javnost—The Public. 2001. Special issue on "The European Public Sphere: Dreams and Realities." Vol. 8, No. 7
Koopmans, Ruud and Jessica Erbe. 2004. Towards a European Public Sphere? Innovation: The European Journal of Social Science Research 17, 2, 97-118.
Koopmans, Ruud and Paul Statham. 2002. The Transformation of Political Mobilisation and Communication in European Public Spheres: A Research Outline. <http://europub.wz-berlin.de>
Le Bart, Christian. 2004. L'invention du lecteur européen. La revue de presse européenne d'Alex Taylor sur France Inter. In D. Marchetti (ed.), En qugte d'Europe. Médias européens et médiatisation de l'Europe, 203-29. Rennes: Presses universitaires de Rennes.
McDonald, D. G. 1990. Media Orientation and Television News Viewing. Journalism Quarterly 67, 11-20.
Page, Benjamin I. 1996. The Mass Media as Political Actors. Political Science & Politics 29, 20-24.
Pfetsch, Barbara, Christiane Eilders and Friedhelm Neidhardt. 2004. Das "Kommentariat": Rolle und Status einer Öffentlichkeitselite. In C. Eilders et al. (eds.), Die Stimme der Medien. Pressekommentare und politische Öffentlichkeit in der Bundesrepublik, 39-73. Wiesbaden: VS Verlag für Sozialwissenschaften.
Ringoot, Roselyne and Jean-Michel Utard. 2004. L'Europe vue par la presse quotidienne
régionale. Les exemples comparés de Ouest-France et des Derničres Nouvelles d'Alsace. In D. Marchetti (ed.), En qugte d'Europe. Médias européens et médiatisation de l'Europe, 24562. Rennes: Presses universitaires de Rennes.
Scherer, Helmut and Simone Vesper. 2004. Was schreiben die Anderen? Ausländische
Pressestimmen als Vorform paneuropäischer Öffentlichkeit - Eine Inhaltsanalyse deutscher Qualitätszeitungen. In L.M. Hagen (ed.), Europäische Union und mediale Öffentlichkeit. Theoretische Perspektiven und Befunde zur Rolle der Medien im europäischen Einigungsprozess, 195-211. Köln: Halem.
Schlesinger, Philip R. and Deirdre Kevin. 2000. Can the EU Become a Sphere of Publics? In E.O. Eriksen et al. (eds.), Democracy in the EU: Integration through Deliberation?, 206-29. London: Routledge.
Schütz, Walter J. 2001. Deutsche Tagespresse 2001. Media-Perspektiven 12.
Sreberny-Mohammadi, Annabelle and Robert L. Stevenson. 1999. Comparative Analysis of International News Flow: An Example of Global Media Monitoring. In K. Nordenstreng et al. (eds.), International News Monitoring, 59-72. Cresskill NJ: Hampton Press.
Stevenson, Robert L. 1998. Remapping the World of the News. In S. Immerfall et al. (eds.), Territoriality in the Globalizing Society: One Place or None?, 111-23. New York: Springer.
Sunstein, Cass R. 2001. Republic.com. Princeton, NJ: Princeton University Press.
Sunstein, Cass R. 2002. The Law of Group Polarization. Journal of Political Philosophy 10, 175-95.
Wu, Haoming Denis. 1998. Investigating the Determinants of International News Flow. A Meta-Analysis. Gazette. The International Journal for Communication Studies 60, 6, 493-512.