Intertextuality in the News : The Spanish Terrorist Attack on March 11

We can define intertextuality as the relationship between one text and other texts. In this paper we analyze intertextuality in news reports focusing upon reporting speech, i.e. the presence of others' words in a text. We argue that when text producers choose to employ reported discourse they pursue a clear objective. Reporters may want, for instance, to detach themselves from what is said or adopt somebody else's words as if they were their own. Thus, intertextuality constitutes a powerful tool at reporters' disposal to suit their own purposes and make texts more persuasive. In the two articles on the Spanish terrorist attack from different newspapers we analyze this is highly significant because after the attack it was not certain who the perpetrators had been. However, authorship was to play a decisive role in the general elections to be held three days later: ETA's hypothesis favoured the political party in power while the fact that Al Qa'ida was behind the attack was beneficial to the opposition. In this paper we show how both newspapers, drawing mainly on the same external voices, succeed in building two different arguments and lead readers towards a biased interpretation of the facts. “The reporters use the voices of the people and groups they report on to tell their story as much as a novelist uses characters or a ventriloquist uses dummies”. (Bazerman 2004: 86) 196 Revista Alicantina de Estudios Ingleses


Introduction
The term intertextuality appeared in the 1960's and since then it has been widely used across different disciplines.Broadly speaking, we can define intertextuality as the relationship that exists between one text and other texts.In our work we will analyze intertextuality in news reports focusing upon reporting speech, i.e. the presence of others' words in a text.We argue that when text producers choose to employ indirect discourse, direct discourse or scare quotes to reproduce what others have said, they pursue a clear objective.Reporters may want, for instance, to detach themselves from what is said or adopt somebody else's words as if they were their own.Thus, intertextuality constitutes a powerful tool at reporters' disposal to suit their own purposes, make texts more persuasive and lead readers towards a biased interpretation of the facts.
In the two articles on the terrorist attack on Spanish trains in 2004 we will analyze this is highly significant because after the attack, it was not certain whether it had been perpetuated by ETA or Al Qa'ida but authorship was to play a decisive role in the general elections to be held three days later: ETA's hypothesis favoured the political party in power while the fact that Al Qa'ida was behind the attack was beneficial to the opposition.

Theoretical Framework
For those involved in discourse analysis texts usually offer a rich series of surface (or textual) marks that can be analysed.However, even though the marks found in texts are generally linked and their interrelation may convey interesting insights, it is always useful to dwell on separate aspects at a time to see how a specific topic is exploited in a given text.In our case the focus of analysis will be intertextuality, described as one of the "analytical focuses which seems especially fruitful in discourse analysis" (Fairclough 1992: 74).In general, authors tend to agree on the fact that intertextuality points to the relation existing between one text and other (previous or future) texts.However, the bibliographic review reveals that intertextuality is a truly flexible concept and that it is possible to find as many definitions of the term as authors have written about it.
Many writers have made reference to this fact in their accounts.Meinhof et al.'s Intertextuality and the Media (2000), which is a clear example of how approaches to intertextuality can vary as individual authors do different textual analyses on media texts (which range from advertisements to music television) within the common frame of intertextuality, describes intertextuality as "a fuzzy yet powerful term" (2000: 1).Bengoechea and Sola (1997), in their introduction to Intertextuality/Intertextualidad, talk about a "chaotic situation" making reference to the fact that many sciences related to knowledge and language have used intertextuality differently.Graham (2000: 2) sees intertextuality as "one of the most commonly used and misused terms in contemporary critical vocabulary".The same author concludes that in order to understand why intertextuality has taken on its current meanings and applications we should study the term's history.
Such a task can be most useful to grasp why (and how) intertextuality, a term which origins are closely related to the literary field, has been so widely studied in various distant fields of activity such as politics, didactics and the film industry or why hypertext computer systems are the last approach to intertextuality.This paper does not intend to give a detailed account of the development of the term intertextuality; however, a brief review will provide the general background needed to get familiar with the topic to be dealt with.
Many books on intertextuality point to Bakhtin as the starting point for the theory of intertextuality and acknowledge that the term was coined by Kristeva in 1967 in her account of Bakhtin's works Rebelais and his World and Problems of Dostoevsky's Poetics.In coining the term Kristeva meant to address a concept radically different from the traditional romantic conception of the literary work in which the text was autonomous and had one identifiable author.In this sense, as Bengoechea states, intertextuality has to do with "el proceso (necesariamente social) de creación de significado".Bakhtin (1986: 69), within the frame of his dialogical theory of language, argues that there are always preceding utterances with which a given utterance "enters into one kind of relation or another (builds on them, polemicizes with them, or simply presumes that they are already known to the listener).Any utterance is a link in a very complexly organized chain of other utterances".As we can see, intertextuality accentuates the dialogue between the different voices that are present within a text.The idea that texts draw upon other texts is also adhered to by Foucault, who states that "there can be no statement that in one way or another does not reactualize others " (1972: 98).Bengoechea, who defines intertextuality in a wide sense as "la relación habida entre un texto y otros que le preceden", divides the various approaches to intertextuality in two different groups: those who analyse texts in relation to the social structures in which they were produced and those who do not.The author acknowledges, however, that the approaches to intertextuality are potentially infinite.
Among those who focus exclusively on the text we can mention De Beaugrande, who, in his work on text linguistics, takes intertextuality as one of the seven principles that make up the concept of textuality.For him, intertextuality involves the relationships with other texts, specially those coming from the same text type or a similar one (De Beaugrande and Dressler 1981).
Authors like Fairclough and Kress are within the second group of analysts, those who consider not only texts but also their social context of production.Intertextual analysis is for them a valuable tool to understand the conflictive socio-cultural representations found in texts.In Discourse and Social Change Fairclough claims that intertextuality constitutes a major focus of analysis because, as Kristeva observes, it implies "the insertion of history (society) into a text and of this text into history".Fairclough explains that by "the insertion of history into a text" Kristeva means that texts are built out of texts from the past and that by "the insertion of a text into history" she is pointing to the fact that texts respond to, reaccentuate and rework past texts.Consequently, texts play a major role in history as well as social and cultural change (Fairclough 1992: 102).
In the same book  develops an analytical framework for text analysis which covers aspects of production and interpretation as well as formal properties of texts.In his view, a comprehensive framework for analysing texts should include: vocabulary, grammar, cohesion, text structure, the 'force' of utterances, the 'coherence' of texts and their 'intertextuality'.The author differentiates between 'manifest intertextuality', the case where specific other texts are overtly drawn upon within a text and 'interdiscursivity', which results in the constitution of texts out of elements and conventions of different discourse types or genres.
In our work we will approach intertextuality concentrating exclusively on what Fairclough calls manifest intertextuality, i.e. features which are manifest on the surface of texts.Within the glossary in Analysing Discourse Fairclough (2003: 39) provides the following definition: The intertextuality of a text is the presence within it of elements of other texts (and therefore potentially other voices than the author's own) which may be related to (dialogued with, assumed, rejected, etc.) in various ways [...] The most common and pervasive form of intertextuality is reported speech (including reported writing and thought), though there are others (including irony).Reported speech may or may not be attributed to specific voices, and speech (writing, thought) can be reported in various forms, including direct (reproduction of actual words used) and indirect report (summary).
Such a definition of intertextuality provides the key elements to analyse texts from an intertextual perspective.According to it, the major focus of analysis should be reported speech (in the form of direct or indirect report).Fairclough suggests, however, that there may be other points or 'forms of intertextuality' worthy of analysis.He considers that irony, presupposition, negation and metadiscourse are all of them forms of manifest intertextuality that can also shed some light onto the intertextual nature of a text.

Intertextuality in news reports
Since our focus of analysis is intertextuality in relation to news reports, we may wonder why text producers resort to intertextuality when they write their reports or, in other words, why reporters choose to include what others have said in their own account of the facts.
In relation to this question, Fairclough (1992: 84) expresses that "intertextuality is basically the property texts have of being full of snatches of other texts […] which the text may assimilate, contradict, ironically echo and so forth".This statement suggests that the presence of others' voices in reporters' texts can serve different purposes.Thus, we can conclude that when text producers mention (directly or indirectly) what someone else has said, they do it for some reason.As Bazerman (2004: 94) clearly states: Intertextuality is not just a matter of which other texts you refer to, but how you use them, what you use them for, and ultimately how you position yourself as a writer to them to make your own statement.People can develop adeptly complex and subtly skilled ways of building on the words of others.Such complex intertextual performances are so familiar we hardly notice them.
Apparently, the use of intertextuality by reporters is always intentional and they can become experts at manipulating others' words to their aim.
Van Dijk's analyses of news are also very illuminating as regards the presence of intertextuality in news reports.In La noticia como discurso (1990: 114) the author examines the journalistic discourse style and concludes that among the limitations news reporters have, due to the fact that they produce institutional discourse, "las citas son una poderosa estrategia para el periodista a fin de evitar las limitaciones sobre la impersonalidad, las opiniones, el punto de vista y la formalidad".His idea also points to the fact that quotations may be used by reporters to suit their own purposes.Van Dijk's studies on news revolve around the structure of news reports as well as the processes of news production and comprehension.He links textual and sociocultural analyses mainly to show how some press practices contribute to the reproduction of racist ideologies.In Racism and the press (1991: 151-175) van Dijk devotes a chapter to quotations and sources.There the emphasis is on the selection and summarization source statements undergo and the ideological basis the strategies employed in those processes may have.Related to this, van Dijk talks about the functions quotations have.We could say that, basically, quotations may serve one of these five purposes: 1. being newsworthy in their own right 2. making the story more lively 3. enhancing the credibility of the account 4. allowing interpretation of news events 5. allowing the insertion of subjecive interpretations and opinions Bazerman (2004: 86) Morawski (1970) and Plett (1991) also coincide in viewing intertextuality as having specific functions within texts.Bazerman considers that others' statements can be used to background, support or contrast arguments; Morawski speaks about the authoritative, erudite and ornamental functions of quotations and Plett holds that given the authority of the original source, authoritative quotations are used to legitimate statements, being their function fundamentally ideological.
As we can see, quotations (and intertextuality as a whole) play an important role within texts and can therefore be a reporter's powerful strategy to make news reports more persuasive.

Terrorism and the media
In the present work we argue that when intertextuality is present within news reports it usually performs a specific function.It is our aim then to analyse the articles we have chosen and find how intertextuality is employed to pass judgement and convey a particular interpretation of the news events reported.
We should mention here that our corpus consists of two news reports on the massive terrorist attack which left near two hundred casualties and more than one thousand injured in Madrid on May 11 2004.The articles appeared on La Vanguardia and El País the day after the event.The significance of the articles chosen lies in the fact that the terrorist attack took place three days before the Spanish general elections were held.Immediately afterwards, it was not certain whether the attack had been perpetuated by Al Qa'ida or ETA but authorship was to play a decisive role in the election results.The ETA-hypothesis favoured the political party in power (the Popular Party), its candidate for presidency and the Prime Minister, J. M. Aznar.On the other hand, in case the Islamic terrorist group was to blame, the opposition (the Spanish Socialist Workers' Party, PSOE) would have had great chances of winning the elections.Eventually the PSOE had a narrow victory which was linked to the prevalence of Al Qa'ida's hypothesis over ETA's.In this context, if we accept that the media has a major role in forming public opinion, the position adopted by the newspapers is to be regarded as highly influential.According to all that has been exposed above, there are good reasons for analysing the intertextuality of a text.
But before we start with the analysis let us first revise some literature on terrorism and the media because the studies done and the experts' opinions can illuminate the analysis of intertextuality in news reports on terrorism.
Most bibliography on terrorism and the media usually revolves around similar topics.Authors (see for instance Aulestia (2005) and Norris et al. (2003)) generally attempt to define terrorism and although their definitions vary, there seems to be wide consensus on terrorism having political causes and political goals (which may be the reason why many people see the media as a mediator between government and terrorists).There is not such agreement, though, as regards the most salient topic, namely, the relationship between the media and terrorism.The nature of the relationship between terrorism and the media is very complex and has led people to take different stances as to whether media coverage on terrorism has positive or negative results.
For some critics terrorists' appearance on the media cannot be negative because the fact that their massacres become news does not mean that reporters support them.Nacos (1994) accepts that the media may benefit terrorists but argues that when they are portrayed negatively or if they are not paid much attention by the media, terrorists become weaker.For authors like Paletz et al. (1992), Rodrigo Alsina (1991) and Soria (1987), however, media coverage of terrorist acts has a negative influence.According to them, terrorists use the media as a means to achieve their goals because the media help them in propagating violence together with the fear, insecurity and chaos they want to transmit.We could say that in their view, without media coverage, there would be no terrorism.M. Thatcher once insisted on this idea stating that ways should be found "to starve the terrorists and hijackers of the oxygen of publicity on which they depend" .The expert Laqueur also adheres to this 1 posture and goes so far as to state that the media are the terrorists 'best friend'.Some authors even talk about a symbiosis or a mutual attraction between terrorism and the media pointing to the fact that the media need great news in the same way terrorists need publicity.In other words, for this group the media and terrorism are mutually interdependent.Finally, there are others who think that all publicity, good or bad, benefits terrorists because all they want is to inflict terror.
Another significant topic dealt with in most books on terrorism and the media is the role of the media in forming public opinion.It is mentioned repeatedly that reporters' work implies selecting and structuring information to present an event into a news form to the public.In this process the media can frame the news in order to guide the audience towards a particular interpretation of the facts presented.This aspect, especially in the press, is very much related to language since it is through the combination of words (though images are important too) how text producers convey their ideas and reflect and reinforce ideologies.Rodrigo Alsina (1991: 19) states that "al analizar [las] estrategias [del discurso periodístico sobre el terrorismo] puede apreciarse lo que pretenden ocultar y lo que desean hacer visible".
If we link the two main topics we have developed (the relationship between terrorism and the media and role of the latter in shaping public opinion) we can see that not only what is reported but also how it is reported matters.In other words, there is more to terrorism than the mere question about whether the media should cover terrorists attacks or not.And that something else, we believe, can be found in language.Analysing news reports can shed some light on the social uses of intertextuality and allow us to see how reporters select and rearrange what others have said to suit their own purposes.

Analysis
We have already described our corpus, which consists of two news reports on the bombing in Madrid on March 11 2004.The articles were selected from the Spanish newspapers El País and La Vanguardia (henceforth referred to as EP and LV) and it should be noticed that the news item from EP is much longer than the one in LV (the former contains 37 paragraphs and the latter 17).For the purposes of our work content matters more than length; however, we will point out those cases where findings may have been influenced by this difference.
We will analyse intertextuality in news reports to see how reporters select and manipulate what others have said to suit their own purposes.As we have previously mentioned, the fact that the attack was perpetuated by Spanish or Islamic terrorist groups played an important part in the results of the general elections to be held three days after the attack.Consequently, we want to see whether reporters use others' speech to support ETA's or AlQa'ida's hypothesis about March 11 terrorist attack.
In order to carry out the analysis we will compare and contrast the selected articles concentrating on different aspects.First the use of reported speech as such will be dealt with.Then we will analyse whose voices are reported to see if there exists some link between who speaks, how his/ her original words are reported (whether they are presented as direct or indirect discourse) and the effect reporters pursue.During the analysis we will also focus upon reporting verbs because their choice is generally said to be significant.

Reported speech
Reported speech, also called discourse representation or intertextual representation, is considered as "the most common and pervasive form of intertextuality" i.e. the most usual way of incorporating others' speech into one's text .There are two main ways to report 2 what others' have said: we can use either direct or indirect speech.The following examples from one of the news items that make up our corpus (the one in EP) will allow us to point out the main characteristics of each type of reported speech: i Pese a todas esas evidencias [desfavorables a la hipótesis de que el atentado era obra de ETA], el ministro del Interior, Ángel Acebes, despejó las dudas en una conferencia de prensa convocada a las 13.00: "ETA ha conseguido su objetivo.El Gobierno no tiene ninguna duda de que ETA esta detrás".ii Interior ha mantenido hasta ahora que la infraestructura de ETA en Madrid era mínima.
At first sight we can see that in i. Acebes' words are explicitly demarcated while in ii.his words (on behalf of the Interior Ministry) are not immediately spotted because they are merged in, i.e. the reporter has incorporated Acebes' words as if they were his/ hers.What makes the intertextuality evident in i. is the use of quotation marks, which render explicit the boundary between the voice of the person being reported and that of the reporter.In direct reported speech we assume that the words are the exact ones the speaker used.Besides, the tense and deictics employed (though we do not find instances of the latter in our example) are those of the original version.
On the contrary, when the text producer resorts to indirect speech to retell what someone else has said quotation marks disappear, tense and deictics change to suit the writer's style and we can never be sure whether the words belong to the original speaker or not.Was la infraestructura de ETA en Madrid es mínima Acebes' original wording or is it the reporter's choice of words?And the final comment: Además, los criminales no avisaron previamente de sus intenciones mortales, como hace habitualmente ETA en sus atentados was also made by the Interior Minister or is it the reporter's?All these questions aroused by the use of indirect reported speech suggest that the voices presented in the news are not clearly demarcated.In this way indirect speech becomes a powerful device at reporters' hands to manipulate the content of the reported text at their will.
Another type of reported speech closely linked to direct reportage are scare quotes i.e. placing single words or short expressions in quotation marks.We could say that scare quotes share in all the characteristics of direct reported speech in that the outside voice is clearly demarcated and they fulfil van Dijk's functions of quotations set forth earlier.
However, there must be some reason why the reporter chooses to include others' complete sentences or mere fragments of what was said.According to Fairclough (1992: 119), scare quotes can be used to distance oneself from the outside voice or use its authority to support one's position.
These two antagonistic uses of scare quotes reinforce the idea that intertextuality can make texts more persuasive because writers can manipulate the content of others' words at their own will to suggest particular interpretations of the facts.
We will analyse now the use of reported speech in each of the chosen articles.We can see that despite the difference in length mentioned earlier (EP article is longer), both articles present an elevated number of cases of reported discourse.This could be linked to the fact that terrorist attacks are a delicate and controversial topic and that is why it is always relevant and wise to listen to what important people have to say about what happened.
Even though both articles share the considerable amount of reported speech they contain, there exists much less similarity in the forms of reported speech reporters employ.It is interesting to notice that while LV makes use of direct speech only once (in order to retell the King's words), EP resorts to direct speech thirteen times.Similarly, the use of scare quotes is higher in EP.It is pretty clear that the reporter writing for LV is more prone to employ indirect rather than direct discourse and thus reproduce others' words merging their meanings with his/ her own.A possible explanation for these findings could be sought in the functions direct quotations may serve.We stated earlier that words said by experts and important people give the writing a sense of factuality and veracity because their content comes from sources that can hardly be challenged.Besides, quotations convey liveliness and turn the news more readily credible.The results make sense, thus, if we link the roles quotations play to the considered newspapers' profiles.LV, being a more conservative newspaper, is likely to employ less direct discourse so as to avoid conveying a sensationalist image full of people's comments.On the contrary, if we see EP as having a socialist profile, it is not surprising that many more people are given a voice in their account of the facts.It is precisely to that aspect, voices, that we turn our attention now.

Reported voices
We have just seen that although we find numerous instances of reported speech both in LV and EP, the former employs much less direct speech and scare quotes than the latter.The use of indirect speech, however, is pretty similar in both articles though we could argue, recalling that LV article is shorter than the one from EP, that indirect speech seems to be a bit overused by LV reporter.
Let us focus now on the voices included because so far we have studied how reported speech was implemented, i.e. we have compared the use of direct speech vs. indirect speech but we still do not know whose words we are talking about.Therefore, we will look at the people who are quoted in each article and we will see how those people's (or institutions') speech is represented in order to see if there exists a link between whose words are retold and they way it is done (by using indirect speech, direct speech or scare quotes).At the same time we will analyse the reporting verbs employed and the place outside voices are given within news reports because we consider these two elements can also help us interpreting the results.We believe that all these aspects will shed some light on the use of intertextuality in news reports: they will allow us to see if, as we believe, intertextuality is a powerful strategy used by reporters to convey particular meanings and lead readers towards biased interpretations of the facts.
The following table shows the different voices that appear in each of the articles as well as the number of times they are included:

Nº of voices: 9 Nº of voices: 22 Table 2. Voices in news reports
The first striking difference between the two reports lies in the number of people whose voices are incorporated.The fact that the article from EP contains twenty-four voices while LV presents only nine is highly noteworthy.If this finding is related to the ones in the previous section, we can conclude that the higher the number of instances of reported speech, the larger the number of external voices included.Again it seems that EP's profile turns its reporter more willing to draw onto others' voices than LV's.
Let us now look at whose voices we are dealing with to discover who gets quoted in each newspaper.We can see in Table 2 that there are six shared external sources (those marked in italics), i.e. six people or institutions whose voices are present in LV and EP in their account of the terrorist attack: the Interior Minister, the King, Al'Qa'ida, Batasuna, the Prime Minister J. M. Aznar and the government.There is a seventh shared source that we decided to call 'anonymous' (because although the comments are present the source they belong to is not indicated).Regarding anonymous sources we can assume that everything that is not attributed to someone else is the reporter's voice.Another possibility is identifying whose words are being included without mentioning who the original speaker was.In that case we should also analyse their content to see if they play any special role.In Table 2 we can also see which are the two voices that appear in LV but not in EP (Audiencia Nacional and the Ecuatorian president) as well as the other fifteen testimonies present in EP (ranging from Bush's to the Pope's) which are absent in LV.Now that we have the gist of the intertextual structure of each article, let us analyse more thoroughly those aspects mentioned at the beginning of this section which can help us confirm our hypothesis.
We think it useful to focus on what we consider the main actors in a terrorist attack.In simple terms, when we think about bombing events first we usually think of its consequences, i.e. the damage caused: the number of victims and the destroyed property as well as how, when and where it happened (thus, we think of experts such as the police, doctors, etc.).Then we wonder who is to blame for the attack (the perpetrators) and finally we expect the government's reaction (because we tend to link terrorism to politics).Consequently, we would expect these three sources (experts, perpetrators and the government) to be present in the account of the facts because they are the main actors we associate with our mental frame of terrorist attacks.

The experts' voice
Let us start with the experts to see whether they are given a voice to inform us of the disaster with accuracy.Both articles provide the reader with the exact number of casualties and injured people together with a thorough description of the event giving exact places and times.However, it is not until we reach the middle of the article in EP or in the very last paragraphs in LV that reporters mention the sources: fuentes de la Audiencia Nacional y de la Delegación del Gobierno in the latter (lines 98-108) and la policía and los expertos (lines 3 91-99) in the former .Not only do we find these sources in non-prominent places in the 4 news report, but we also listen to their voices indirectly reported.What is more, in EP the experts' and police statements lead us to believe that it was Al Qa'ida (and not ETA) which was to be held responsible for the attack.
There are other two interesting passages in EP.In the first one, expressions apparently said by an expert appear in the form of scare quotes.Obviously, to say that there were 44 heridos 'en estado crítico' y otros 27 'muy graves' (line 37) entails accuracy and veracity.We wonder, however, if the common reader is able to tell the difference between being 'en estado crítico' or 'muy grave'.The second passage that caught our attention is one where we can listen to the shocking and moving exact words of a survivor and a nurse (lines 106-109).The visual and auditive images they depict: "miré atrás y era como estar en la guerra", "vi cadáveres con los móviles sonando, muchos jóvenes y niños..".make us feel as if we were there and, as van Dijk explains, "make the story more lively".
On the whole we consider that the experts" voices pervade both articles and the reporters draw on their descriptions, statements and explanations to write down their own account.In disasters like this, we believe, the reader assumes that what is written depicts the facts, i.e. readers do not need the reporter to explicitly say "this is what the police said" or "this is the explanation the bombing expert gave" because serious reporting generally report on what those kinds of sources state.However, it is interesting to notice how reporters play upon this 'common knowledge': at times official sources are backgrounded because it is unnecessary to mention them since it is presupposed that what is said springs from them.
Some other times, however, emphasis is put o the fact that certain piece of information comes from an expert so as to convince the reader that what the writer is saying is completely true.In such a game, intertextuality plays an important part and our findings reflect how the use of the different forms of reported speech are manipulated (especially by EP reporter) to convey not only a feeling of veracity and liveliness in his/her account but also a particular reading and interpretation as to whom the attack must be attributed to.

The terrorists' voice
We have just analysed the experts' voices treatment in the account of the Spanish March 11 terrorist attack.They were considered first because we believe that experts are the ones in a position to inform us about what happened.We should recall, however, that in this attack two main hypotheses about who the perpetrators had been circulated: some facts indicated it had been ETA while others pointed to Al Qa'ida.Thus, the next actors we should consider are the terrorists themselves in order to see if they are given a voice, be it to defend or inculpate themselves.
The two articles coincide in including the terrorists' words.In the case of Al Qa'ida (the second most cited voice in both articles -see Table 2), even though both newspapers include their statements in important places (summaries and first paragraphs), the treatment they give is quite different.In LV we can read that Al Qa'ida "se atribuye la masacre" y "reivindica los atentados" making reference to an e-mail sent to the British newspaper by the Islamic group but that is all we get of the terrorists' voice.The whole passage (where two instances of intertextuality can be found) conveys a feeling of vagueness due to the hypothetical expressions "habrían consumado" and "se especulaba incluso con la posibilidad".Such findings point to the fact that, according to LV, it was ETA (and not Al Qa'ida) who was behind the attacks.
In EP, on the contrary, everything leads us to believe that the perpetrator of the attack was Al Qa'ida.In paragraph three we can read the exact words the Islamic terrorist group sent to the British newspaper and how the hypothesis is supported by successive (antiterrorists' and Interior Minister's) statements.
These results may lead us to think that if LV favours ETA's hypothesis, its article will enhance ETA's voice in the same wat it silenced (through vagueness) Al Qa'ida's.Surprinsingly, when Batasuna (ETA's presumed political wing) appears in the news, not only does it condemn the attack but it also denies ETA's authorship.The strategy employed seems clear: the statements are (shortly and indirectly) quoted as if they had the least importance and they are found in the middle of a passage describing the destruction caused by the attack (lines 46-47).As we interpret it, the reporter writing for LV does not need to say that they support ETA's hypothesis: their strategy lies precisely in not pointing to ETA as the responsible for the attack.
Unlike LV, EP is more explicit.It presents Batasuna's voice (lines 179-182) towards the end of the article and makes it very clear that they think alike.That is, while all the people are debating whether it was ETA or Al Qa'ida, Batasuna confirms that ETA is not to blame (thus, pointing to Al Qa'ida).Like them, EP does not doubt about ETA's authorship because EP knows ETA was not behind the attack.

The government's voice
We will focus now on the third actor we consider essential in a report on a terrorist attack: the government.According to what scholars say about terrorism having political causes and goals and in the light of the political context of the Spanish terrorist attack, it is quite undeniable that the government's voice has a major say in the matter.
We have seen in Table 2 that the Interior Minister A. Acebes (and the Interior Ministry) are soundly convinced that it was ETA.However, while LV destroys all arguments that attempt against ETA's hypothesis, EP shows a more doubtful Acebes (in order to suggest it was Al Qa'ida).The writers achieve this through a witty argumentation pattern and the manipulation of reporting verbs and forms of reported speech employed (indirect speech, direct speech or scare quotes).
As regards argumentation, in LV the use of the discourse marker pero (line 4) is highly significant because it counterargues what has been said before emphasising what comes afterwards.In lines 22 to 26 there are no connectors after the long indirect speech reportage but the short line (another instance of indirect speech) at the end of the paragraph undermines the previous argument making it clear that there are no reasons to think that Al Qa'ida could be held responsible because, according to Acebes, "el texto no contenía ninguna amenaza".The reporting verbs 'se aseguró' and 'consideró sin titubeos' (lines 22 and 62) emphasise Acebes' words and give stronger support to ETA's hypothesis because they seem unchallengeable.
In EP Acebes' voice also blames ETA, but its deep conviction vanishes as the expressions "expresó algunas dudas", "admitió algunas dudas" and "tampoco confirmó" (lines 29, 55 and 57) used to contextualise his comments cast doubts here and there.In fact the whole passage (lines 44 to 58) reveal how Acebes is shown assertive but not very convincing at the same time as he confirms it was ETA but some hours later (as more evidence is gathered) he cannot be so sure any longer.
In the fifth paragraph it seems logical that EP reporter chose to represent Acebes's words through direct quoting.In this way, the text producer is detaching him/herself from what was said revealing that EP thinks the opposite.This can be contrasted with the use of indirect reported speech made by LV when reporting Acebes's words.We showed earlier how this form of reporting speech implied the merging of voices and this is what LV aims at so as to show that they also support ETA's hypothesis.
Before concluding this section about the government's voice let us recall that the terrorist attack took place three days before the Spanish general elections were held and though it was not certain whether the perpetrators had been Al Qa'ida or ETA, authorship was to play a decisive role in the election results.The reason for this was that if Al Qa'ida was to blame, the Popular Party (PP) would have been doomed to lose the elections because the attack would have been seen as a reprimand for Spain's participation in the war with Iraq.That is why the PP favoured ETA hypothesis while the Spanish Socialist Workers' Party (PSOE) pointed to Al Qa'ida in order to persuade people to vote for them promising that they would withdraw the troops and would no longer support the war.
In this context we consider it relevant to consider other political voices present in the articles.The Prime Minister at the time of the attack, J. M. Aznar, is present once in each of the news reports and although he is expected to defend the PP (and its foreign policy) and hold ETA responsible for the attacks, we find that he does not mention this terrorist group.On the contrary, his words are general and include the terms 'terrorists' and 'assassins'.Both newspapers emphasise the fact that he did not mentioned ETA at all in his speech (line 61 and 68 in LV and line 71 in EP).On the one hand, this fact suits EP (not mentioning ETA in a way supports Al Qa'ida's hypothesis) and, on the other, helps LV to build their argument: they do not need to say it explicitly because it is all very clear, at least for them, that it was ETA the responsible for the attack.We suggest reading some lines further to see how the interplay of important voices in LV builds the argument in this way.
Other political voices can be found in the candidates for presidency for the PP, M. Rajoy and the PSOE, J. L. Rodríguez Zapatero.While both are absent in LV, we can hear their words in EP.The first time EP writer presents them together (line 6) (as if there were no rivalry between them) asking for the democrats' unity in an attempt to scorn the terrorists' intention to confront the Spanish people.Later, Zapatero's words are quoted verbatim and when repeating his pledge he asks Spaniards to cast their vote "como reacción a ETA" (lines 142-45).It seems surprising that Zapatero himself attributes the massacre to ETA and not to Al Qa'ida (which would be beneficial for him in the elections), but it can be part of his strategy or the mere fact that it was not till March 13 that Al Qa'ida's hypothesis gained weight.After Zapatero's statement EP presents three more arguments supporting ETA's hypothesis: Ibarretxe's, Llamazares' and Negroponte's (lines 154-78), but they occupy unimportant positions in the article and they are quoted either directly or with scare quotes.This, we believe, indicates that the text producer is detaching him/herself from what is said because he/she does not agree with them.
On the whole, we have seen how political voices pervade both reports on the Spanish terrorist attack.The Interior Minister's voice is not only the most cited one, but also the voice which is mainly used by reporters to build their own argument and persuade the reader that it was ETA or Al Qa'ida the terrorist group to be held responsible for the attack.

Other voices
So far we have analysed the intertextual treatment of those voices that in our belief where the most related to the terrorist attack.There remain others, however, that should also be mentioned due to their prominent sources.We refer to King Juan Carlos's and the Pope's words.The former is a salient voice that appears three times in both articles to strongly condemn the attack.The words taken from his speech are cited in the form of scared quotes or direct quoting in both newspapers as if to show that they completely agree with him and accept his words.
There is one last coincidence: according to EP, King Juan Carlos gave a speech "en el que no citó a la banda terrorista ETA" (line 146) while LV states that "tampoco el Rey la citó" (line 62).Taking our previous findings into account, we see once again how this reinforces the intertextual pattern employed in each article.While LV is convinced that ETA is the responsible (and therefore, there is no need to mention it), EP implies that if the King does not mention ETA, it should be because it was Al Qa'ida.
The Pope's testimony (in direct speech) only appears in EP (lines 168-72), a fact which also reinforces this newspaper social side and its willingness to include all those voices that have something to say about what is being reported.

Conclusion
In this work we have analysed intertextuality in news reports in order to show how reporters include what others have said to suit their own purposes an suggest biased interpretations of the facts.Reported speech, the most common form of intertextuality, allows text producers to retell others' words in different forms: indirect speech, direct speech or scare quotes.We have seen that these kinds of reportage have different (and at times opposing) functions because they may allow writers to detach themselves from what is said, support the quoted words or even merge them in as if they were their own.In the articles about the Spanish terrorist attack on March 11 we have analysed we found that intertextuality plays a major role in both in LV and EP and becomes a powerful tool at reporters' disposal to make their texts more persuasive and lead readers towards biased readings.
Our findings showed that both articles contain a high amount of reported speech but they differ in the form of reported speech that prevails.The article in LV makes extensive use of indirect speech (which allows the original meaning to be reproduced according to the reporter's intention) while in EP's account many more instances of direct speech and scare quotes (with various functions) are found.
In both reports we can hear the voices of the three main actors of a terrorist attack: the experts, the government and the terrorists themselves.In all the cases we showed how these sources' statements are incorporated and manipulated at the text producer's will to convey particular interpretations.The writers' pursued effect is achieved through a combination of the form of reported speech and reporting verb chosen together with the prominent or backgrounded place the voices are assigned.
Finally, the argument pattern also plays an important role and helps writers to work on others' statements and fulfil their aims.In this way both newspapers, drawing mainly on the same external voices, succeed in building two different arguments: while LV puts the blame of the bombing on ETA, EP holds that Al Qa'ida is behind the attack.This suggests that reporting is not always as factual as it seems and that, at least in the context of the articles we dealt with (where general elections were three days ahead), political ideologies play a very important role because, ultimately, they are 'expressed' through words.That is why it would be interesting to analyse ideologies in depth and contrast newspapers with marked political orientations so as to see if our findings still hold true.MADRID.-Un miércoles convertido en domingo.Madrid vacío, sólo roto por decenas de sirenas de ambulancia, con un aire de apocalipsis.Un balance de 192 muertos y 1.430 heridos al cierre de esta edición.

Table 1 .
Table1shows the use of reported speech in the articles from LV and EP and the number of instances of scare quotes, direct discourse and indirect discourse found in each article.Use of reported speech.