THE EFFECT OF GENRE ON REPORTING SPEECH : CONVERSATIONS AND NEWSPAPER

This article focuses on reported speech in two different genres: spoken conversation and newspaper articles. There are two basic structures that allow language users to report formerly uttered words: direct and indirect speech. Both structures serve to integrate former discourse into the ongoing discourse. In different genres, however, language users draw upon different language tools to meet their communicative aims. This study examines how this might affect the distribution of direct and indirect reports across conversations and newspaper articles. Two of various hypotheses that have been suggested for the different uses of direct and indirect reported speech are examined using qualitative and quantitative analyses: 1) that direct speech would be a ‘less complex’ strategy than indirect speech, in the sense that the reporter does not need to make deictic adaptations if we take the ‘original’ words as our starting point, and 2) that direct speech is a more involving strategy than indirect speech. While the statistical results confirm both hypotheses, the confirmation of the complexity hypothesis differs for the two genres studied and needs some refinement as will be show with further qualitative analyses.


Introduction
Reported speech (and reported writing) is a phenomenon that has been studied extensively and from multiple perspectives.Some research has focused mostly on the * I would like to thank the two anonymous reviewers and the editor for their helpful comments.form of reported speech and the transformation from direct to indirect speech (Leech and Short 1981, Wierzbicka 1974, Coulmas 1986).This research takes as a point of departure the difference between de dicto and de re.Various studies however have rejected the idea that direct speech would be a literal repetition of the original utterance and indeed that there necessarily is an original utterance -as with reported speech attributed to animals or that is hypothetical (e.g.Tannen 1989, Clark en Gerrig 1990)-.Another line of research analyzes reported speech in context (e.g Bakhtin 1981, Voloshinov 1971, Tannen 1989, Baynham en Slembrouck 1999;Holt & Clift 2007) focusing on how direct speech is foremost a reconstruction or recreation of earlier hypothetical utterances or, in the case of reported writing, the quotes are (almost) literal but still are recontextualized (e.g.van der Houwen under review, van der Houwen & Sneijder submitted).More than a grammatical analysis, these studies emphasize how reported utterances are recontextualized and the different functions that direct and indirect reported speech fulfill in the context in which they occur.
The main difference between a direct and an indirect report is the point of view from which the original words are interpreted.With direct speech, the 'original' utterance is interpreted from the point of view of the reported speaker, whereas with indirect speech the original words are interpreted from the point of view of the reporter.Various hypotheses have been proposed to explain the differences between the direct and indirect report strategies, for example, differences in involvement (e.g.Chafe 1982;Tannen 1989), differences in complexity (Li 1986), differences in (apparent) authenticity/accuracy/reliability (e.g.Holt 1996;Li 1986;Philips 1986;Galatolo 2007).
In this paper, the emphasis is on both form and function but with an emphasis on form.It focuses on two of the hypotheses mentioned: 1) that direct speech would be a 'less complex' strategy than indirect speech (proposed by Li 1986), in the sense that the reporter does not need to make deictic adaptations if we take the 'original' words as our starting point (and whether that is a logical point of departure considering what we now from studies such as Tannen 1989;Clark & Gerrig 1990 will also be further discussed) and, 2) that direct speech is a more involving strategy than indirect speech (proposed by e.g.Chafe 1982;Tannen 1989).To test these hypotheses, 2 different genres are examined: a relatively flexible genre, namely conversations in the sociolinguistic interview and a relatively fixed genre, newspaper articles.In different genres, language users draw upon different language tools to meet their communicative aims (Bhatia 1993(Bhatia , 2005)).While in the sociolinguistic interview the aim is to have a 'natural conversation' where the informants tell about themselves and their situation, newspaper articles aim to inform readers of what is happening in the world.By performing quantitative and qualitative analyses, this article examines whether genre, that is, the communicative goals of a text, affects the distribution of direct and indirect reports. 1he paper is structured as follows.In section 2, direct and indirect reported speech and their deictic nature are examined.In section 3, the two different genres that this study investigates are discussed and an overview is given of the various forms of reported speech found in newspapers and conversations respectively.After a description of the materials and method, the results of the quantitative and qualitative analyses are presented in section 5, followed by the conclusion.

Reported speech and deictic orientation
Generally, two kinds of reported speech are distinguished: indirect and direct report.An example of each is given in (1a) and (1b), respectively.Both forms come from the same speaker, who reports her husband's words twice, once indirectly and once directly.
(1) (a) Me dijo Juan que vio a mi mamá en el cielo.Figúrate que se murió mi mamá tres días antes de su hijo.Y que estábamos rezando por la noche los... rosarios, y cuando salíamos, que... habíamos terminado de los rosarios, me dice en el corredor... me puso la mano en el hombro, (b) y me dice: "Licha, está tu mamá lindísima, gloriosa, en el cielo."(Centro de Lingüística 1971: 143) (a) 'Juan said to me that he had seen my mother in heaven.Imagine that my mother died three days before her son.And at night we were telling our... beads, and when we left, that... we had stopped telling the beads, he says to me in the corridor... he put his hand on my shoulder, (b) and he says to me: "Licha, your blessed mother, is in heaven.' The reported utterances, vio a mi mamá en el cielo 'he had seen my mother in heaven' and "Licha, está tu mamá lindísima, gloriosa, en el cielo."'"Licha, your blessed mother, is in heaven."'contain very similar information, namely, that the mother of the reporter is in heaven.But they differ in the point of view the reporter adopts.In the indirect report (1a), the reporter reports from her own point of view, referring to her husband with the third person pronoun 'he' and to her mother as 'my' mother: he had seen my mother in heaven.In the direct report, the reporter adopts the point of view of the reported speaker, and plays the role of her husband as can be seen from the vocative Licha (the informant) and the possessive pronoun 'you': "Licha, your blessed mother is in heaven" (see also Van der Houwen 1998, 2001) This shift in point of reference, demonstrated most clearly in the above example by the possessive pronouns in mi mamá 'my mother' and tu mamá 'your mother', is one of the characteristics that differentiates indirect and direct speech.Apart from pronouns, other elements may need to be adapted to the new context in which the words are uttered.Section 2.1 explains these deictic changes in more detail.
In general, the occurrences of reported speech found are reports of utterances that were actually, although not necessarily literally, produced in the past.But direct and indirect speech are strategies that also allow speakers to report thoughts, for example, or what could have been said, what will be said or, for that matter, what cannot be said.In (2) I give an example, taken from Tannen (1989: 118), of the latter, where a speaker 'reports' the words of a cat: (2) A guest notices the hosts' cat sitting on the window sill and addresses a question to the cat: "What do you see out there, kitty?"The host answers for the cat: She says, "I see a beautiful world just waiting for me." The label 'reported speech' may thus not be the most appropriate.Tannen (1986: 311) suggests the label 'constructed dialogue': Examination of the lines of dialogue represented in storytelling or conversation, and consideration of the powers of human memory, indicate that most of those lines were probably not actually spoken.What is commonly referred to as reported speech or direct quotation in conversation is constructed dialogue, just as surely as is the dialogue created by fiction writers and playwrights.
Although she only refers to direct speech, the same could be said for indirect speech.Example (3) illustrates an indirect report, where the speaker 'reports' what nobody can say: (3) En ese sentido, reconoció que en nuestro país aún existen muchos millones de mexicanos que padecen de pobreza, injusticia y desigualdad, pero a pesar de esos grandes problemas, nadie puede decir que en México el Siglo XX haya pasado en vano.(EL Sol Jan02prim1) 2 'In this sense he recognized that there still exist in our country millions of Mexicans that suffer poverty, injustice and inequality, but despite those big problems, nobody can say that in Mexico the twentieth century has passed in vain.' While the label "reported speech" might not be the most accurate, I will use it throughout this paper because the term is so widely used.
It is important to state though that rather than seeing reported speech as reporting a 'source utterance' I suggest it is about reporting 'source content.'The 'source utterance' is generally no longer accessible if indeed it was uttered.Studies in psychology have shown that our memory is not in fact such that we would be able to literally report prior speech (Lehrer 1989).So the source utterance is more likely an idea/ some content, rather than an actual utterance, hence 'source content'.This view further implies the deictic adjustments can also be seen as deictic choices the reporter needs to make when reporting this 'source content' either directly or indirectly; the choices have to do with adjusting and recontextualizing the source content to its new communicative context.
The different kinds of adjustments that might be needed to report 'source content' 'can be divided into four categories, two related to deictic orientation and two related to the surface structure: 1) alternation regarding deictic elements such as pronouns, 2 As will be further detailed in section 4, all data from the written corpus come from 'El Sol de México'.The articles come from the first week of January 2000.The source after each of the examples is indicated as follows: 'Jan' stands for January; 02 stands for the date; 'prim' stands for frontpage and the number that follows is the number of the article.(El Sol Jan02prim1) means that it was the second article on the frontpage of 'El Sol de México' on January 2. demonstratives, adverbs of time and place, 2) de dicto-de re encoding, 3) quotation marks or intonation, for a direct report in writing and speaking respectively, and 4) a subordinating conjunction que 'that' or si 'if' may introduce indirect speech.The first two are discussed in the sections, 2.1.and 2.2.The surface features are discussed in section 3.

Deictic elements
In the present paper, the label deictic element is used to refer to those elements that are 'indexes'; what Lyons calls "'orientational' features of language which are relative to the time and place of utterance" (quoted in Leonard 1995: 283).
As shown in (1), the point of view adopted by the speaker is crucial regarding the use of pronouns and possessives.Unless the reported speaker and the reporter are referring to the same person(s) as in: "I i always walk the dog" -I/you/they said that I i always walk the dog, or "They i always walk the dog" -I /you/he said that they i always walk the dog, for example, a pronominal change or some description, in the second case, of who they are is necessary.For the purposes of the present paper I will only look at whether there is a pronominal difference between a direct and a indirect report or not (for a more detailed account on referentiallity of personal pronouns in direct and indirect speech see Li 1986: 30-34).
Whereas the form of all deictic elements depends on real world referentiallity, only adverbs of time and place, and demonstratives depend directly on context.It might be useful when discussing deictics and reported speech to distinguish between the 'original communicative setting' and the 'ongoing communicative setting': pronouns and possessives do not, normally, depend on time or space differences between the original and ongoing communicative setting, while adverbs of time and place, and demonstratives do.Example (4) shows that if the original and ongoing communicative setting overlap regarding time reference no change is necessary when changing the direct report to an indirect report.
(4) Sobre la transición informática al año 2000, puntualizó que los reportes desde el 31 de diciembre hasta este martes son positivos, en virtud de que los sistemas de los sectores estratégicos del país funcionan con normalidad, tal como ocurrió en la mayor parte del mundo.(El Sol Jan05prim2) 'Regarding the information transition to the year 2000, he specified that the reports since the 31st of December till this Tuesday are positive, and given that the systems of the strategic sectors of the country function normally, as occurred in the biggest part of the world.' The utterance in (4) is reported indirectly and contains a time reference: este martes 'this Tuesday'.In this particular example, reported on the Wednesday following the Tuesday referred to, the original and ongoing communicative settings coincide regarding this deictic element.If, however, one would report this utterance a week later, este martes 'this Tuesday' must change to, for example, el martes pasado 'last Tuesday'.This means that if the reporter and the reported speaker have an element of the communicative setting in common, that is, if there is an overlap between the original and the ongoing communicative setting, no adaptation needs to be made regarding this element.To give another example, if the adverb of place here refers to in this pub, in this city or in this country and the speaker, is, at the moment of reporting, in this pub, this city or this country, the reporter does not have to make a deictic adjustment when reporting the content indirectly.
Apparently, the pronominal change is in many languages the only criterion for distinguishing a direct report from an indirect report (Li 1986: 32).According to Li, pronominal change is a universal characteristic of those languages that have both direct and indirect speech as report strategies.Not all languages have both strategies, but a language that has indirect speech necessarily has direct speech.
Tense is another deictic element that adjusts to the reporting context.De Jonge (1995: 20-21) gives an overview of the sequences of tenses in Spanish, indicating how verb tenses from an original utterance need to change if reported indirectly.For example, when the report verb is in the Preterite, the quoted utterance cannot contain a Present or Future tense.Instead, these originally Present and Future tenses would need to be converted to their past equivalent, the Imperfect and the Conditional, respectively.Example (5) shows, however, that even with a report verb in the Preterite, Present and Future verb tenses are possible and quite common indeed in certain types of discourse. (5) Fue entonces cuando Juan Ramón de la Fuente aclaró que la reforma mencionada de ninguna manera contempla la privatización de la Universidad Nacional Autónoma de México, y más bien saldrá un nuevo modelo de universidad pública que el país necesita para el próximo milenio.(El Sol Jan02prim4) 'It was then that Juan Ramón de la Fuente clarified that the mentioned reform, in no way contemplates the privatisation of the Universidad Nacional Autónoma de México, and more likely a new model of the public university will come out that the country needs for the next millennium.' In example (5), the quoted utterance contains two Present tenses and one Future tense.If one were to report Juan Ramón de la Fuente in, say, ten years, the sequencing of tenses, as indicated by de Jonge, would apply.The context will have changed so that the reporter and the reported speaker do not have context in common regarding this particular utterance.In ten years we would know what has happened to the university and the utterance would have lost its 'actuality'.This will need to be reflected in the verb tenses.As with the adverbs of time and place, and the demonstratives, changes in tense are only needed if indeed the original and ongoing communicative settings differ with regard to time reference.

De dicto-de re analysis
Related to the deictic orientation of reported speech is the coding of the direct and indirect reported utterance in terms of de dicto / de re coding.Coulmas (1986) gives an account of the different coding of the report strategies a) the de re coding, which allows the reporter to paraphrase the original utterance and/or add information which may lead to inferences not intended by the original speaker and b) the de dicto coding, where the original utterance remains intact except for the necessary deictic shifts.The classification of de re and de dicto readings of indirect speech brings Coulmas (1986: 2-3) to the following descriptions of direct and indirect speech: Direct speech, in a manner of speaking, is not the reporter's speech, but remains the reported speaker's speech whose role is played by the reporter.[...] In indirect speech, the reporter is free to introduce information about the reported speech event from his point of view and on the basis of his knowledge about the world, as he does not purport to give the actual words that were uttered by the original speaker(s) or that his report is restricted to what was actually said.
According to Coulmas, when reporting indirectly, there are two types of changes: 1) deictic changes (pronouns, adverbs of time and place, tense, etc.); and 2) there are some optional changes depending on whether the speaker wants to report de dicto or de re.While direct speech always purports to be de dicto, indirect speech may be de dicto or de re.If the speaker only makes the necessary deictic changes when reporting indirectly it is considered as a de dicto coding of the original utterance.If, however, the speaker adds information, the original utterance is coded de re.Example (6), taken from Coulmas, illustrates how information may have been added and how an utterance may have a de re or de dicto interpretation: (6) John says that this lunatic doesn't know what he is saying.
According to Coulmas (1986:4), this lunatic in (6) may be part of the report or an addition by the reporter, in which case the utterance is coded de re.If, however, this lunatic was actually part of the original utterance the coding is de dicto.
The de dicto-de re classification is interesting, but hard to apply.Since one, generally, does not have access to the 'original utterances', the coding of an indirect report will generally be ambiguous which makes the use of such classification rather doubtful.The implication is, however, that reporting directly the reporter gives the interlocutor the impression that those words were the exact same words as used by the reported speaker; reporting indirectly the reporter does not give such impression.This supports the second hypothesis that will be tested in this article, namely, that direct speech is a more involving strategy than indirect speech because pretending that one reports the original words will involve the interlocutor more than if the interlocutor thinks it is a paraphrase.Section 3.4 explains the hypothesis in more detail.

Genre: surface differences when reporting in newspapers and conversation
In spoken language, probably the most important feature that makes the distinction between direct and indirect speech is intonation.But in written language there is no intonation to rely on.In the present study, since the analysis of the spoken corpus is based on transcriptions of the conversations it also relies on written language in the sense that, instead of intonation, quotation marks indicate the occurrence of a direct report.The following is, therefore, an account of the different forms of reported speech found in the conversations and newspaper articles analyzed.

Conjunction que
As described in section 2, direct and indirect report may be distinguished based on their deictic differences.But direct and indirect speech are also distinguished based on what grammar prescribes and what we learned in school.For instance, the only reason why we interpret (4), above, as an indirect report, is the presence of the conjunction que 'that' and the absence of quotation marks.Thus, without changing the words themselves we can convert the utterance into a direct report as presented in ( 7): (7) Fue entonces cuando Juan Ramón de la Fuente aclaró: "La reforma mencionada de ninguna manera contempla la privatización de la Universidad Nacional Autónoma de México, y más bien saldrá un nuevo modelo de universidad pública que el país necesita para el próximo milenio." 'It was then that Juan Ramón de la Fuente clarified: "The mentioned reform in no way contemplates the privatisation of the Universidad Nacional Autónoma de México, and more likely a new model of the public university will come out that the country needs for the next millennium."'This would mean, then, that if no deictic adjustments are needed, we rely on the conjunction que 'that' or quotation marks (or intonation in spoken language) in order to interpret a quote as direct or indirect.
Examining the different ways in which utterances are reported, there does not seem to be such a clear linguistic difference between a direct and an indirect report.First, as is shown in ( 4) and ( 5), if the reporter's perspective regarding time, space and people is identical or at least very similar to that of the reported speaker's, no deictic adjustments are required if reported indirectly.Second, not all indirect reports ask for the conjunction que 'that'.An example of an indirect report with a different format, not the more frequent [report verb + que + quote], but a quote interrupted by the report verb [quote + report verb + quote] is (8a).In (8a), the italicized prepositional phrase con su cliente 'with their client', containing a possessive adjective su 'their', indicates that it is indeed an indirect report.A direct report of the same utterance would have been with the possessive adjective nuestro 'our' as follows: (8) a.El hecho de hallarse en prisión, indicaron los abogados de Cabal en la petición de libertad, dificulta la organización de la estrategia de defensa con su cliente y añadieron que no hay ninguna necesidad de mantenerlo encarcelado, pues no ha escapado de la cárcel en ninguna ocasión anterior.(El Sol Jan05prim3) 'The fact that he is in prison, indicated the lawyers of Cabal in the plead for freedom, makes the organization of the defense strategy with their client difficult and they added that there is no need to keep him in prison since he never has escaped from prison in any former occasion.'b.El hecho de hallarse en prisión, indicaron los abogados de Cabal en la petición de libertad, dificulta la organización de la estrategia de defensa con nuestro cliente y añadieron que no hay ninguna necesidad de mantenerlo encarcelado, pues no ha escapado de la cárcel en ninguna ocasión anterior.
'The fact that he is in prison, indicated the lawyers of Cabal in the plead for freedom, makes the organization of the defense strategy with our client difficult and they added that there is no need to keep him in prison since he never has escaped from prison in any former occasion.' Similarly, the pattern [report verb + que + quote] is not exclusive of indirect reports.To complicate things further, some direct quotes are introduced by que 'that', making this conjunction an element that may co-occur with both indirect and direct speech, as shown in the following example: (9) (a)Y al dar respuesta a las críticas de sus opositores sobre su propuesta en esa materia [corrupción en México], el candidato presidencial del PRI señaló que "están concentrados más en las críticas que en hacer propuestas que ofrezcan soluciones a los problemas del país''.(b) Subrayó que "lo que les preocupa es llegar a Los Pinos, están preocupados en tener un lugar donde vivir, un coche que conducir, un avión en el que viajar, pero no hacen propuestas, porque no las tienen, para superar lo que nosotros hemos presentado''.(El Sol Jan06prim5) (a)'And while answering the criticism of his opponents on his proposal regarding that matter [corruption in Mexico], the presidential candidate of the PRI indicated that "they are more concentrated in criticizing than in making proposals that may offer solutions to the problems of the country".(b) He underlined that "what concerns them is to arrive at Los Pinos 'the white house', they are preoccupied with having a place to live, a car to drive, a plane to travel with, but they do not make proposals, because they do not have them, to do better than what we have presented".' Example (9) contains two examples of direct speech introduced by a report verb and the conjunction que 'that' In (9a), there is no difference with an indirect report except for the quotation marks.In the second instance of reported speech, (9b), however, it is clear that it is indeed a direct quote introduced by a report verb and que 'that'.In this example, para superar lo que nosotros hemos presentado 'to do better than what we have presented' the deictic orientation of nosotros 'we' indicates that it is a direct report.If reported indirectly, this pronoun would have been ellos 'they': Para superar lo que ellos han presentado 'to do better than what they have presented'.A report with the pattern [report verb + que + quote] was found only once in the conversations studied; it is given in ( 10): (10) Ese aliento es la vida, ese aliento es Dios.Para nuestros fines vamos a llamarle la vida, para los fines de nuestra plática.Hay un axioma esotérico que dice que "como es arriba es abajo, como es abajo es arriba".(Centro de Lingüística 1971:112) 'That aliment is life, that aliment is God.For our purposes, we are going to call it life, for the purposes of our conversation.There is an esoteric axiom that says that: "As it is above it is below, as it is below it is above".' In (10), the transcriber must have been led to put quotation marks either by the intonational features provided by the speaker, or by the fact that the 'reported speaker' is not a human being but an axiom and, therefore, the transcriber puts the quotation marks as one would do when quoting from a book.

Punctuation
Although, generally, punctuation in written language plays an important role in guiding the reader as to how to interpret what s/he is reading, this is not always the case.The following fragment gives an example of an indirect report changing into a direct report without any punctuation indicating such change.
(11) [from an article about child theft; five Mexican children were found in Canada] Añadió Estrada que tiene el conocimiento que el gobierno canadiense ha manifestado la voluntad de naturalizarlos [los niños], hecho del que no estamos en contra, pero deseamos verificar en todas las procuradurías de los estados mexicanos, para establecer si alguno de ellos se encuentra reportado como robado.(El Sol Jan06prim1) 'Estrada added that he knows that the Canadian government has indicated that it is willing to naturalize them [the children], something we are not against, but we want to verify with all the lawyer's offices in the Mexican states, to see whether one of them is being reported as stolen.' In example (11) the reporter begins reporting from his own point of view, tiene el conocimiento 'he knows'.Within the same sentence the reporter changes the deictic orientation and adopts the point of view of the reported speaker as follows from the verb conjugations estamos 'we are' and deseamos 'we want'.From context it is unlikely that the reporter is including him/herself.The reporter, after all, is only reporting and is not taking part in the negotiations regarding the return of these robbed children that are now in Canada.3Examples like (11), where there is no punctuation indicating a change in deictic orientation, are not common, as may be expected from their ambiguity.However, the mixing of indirect and direct speech (as indicated by quotation marks) is not uncommon in newspapers (18.5% of the occurrences in the newspaper articles analyzed).The following fragment gives an example of such mixing.
(12) El juez Alan Goldberg indicó que la historia de Cabal Peniche, que huyó de México en 1994 y viajó de un país a otro para escapar de la justicia hasta que fue descubierto por la policía en Australia, "me lleva a rechazar la presunción de que el riesgo de fuga del solicitante es pequeño".(El Sol Jan05prim3) 'Judge Alan Goldberg indicated that the story of Cabal Peniche, who fled from Mexico in 1994 and who traveled from one country to another to escape from the law until he was discovered by the police in Australia, "leads me to reject the presumption that the risk of fleeing is small".' In example ( 12) the report is introduced as an indirect report, with an embedded direct report as indicated by the quotation marks and, more importantly, by a shift in deictic orientation: "me lleva .." '"Leads me .." ', the pronoun me 'me' indicates that it is a direct report.An indirect report would have, the third person pronoun lo 'him'.
As the examples have shown, there is not a clear-cut distinction between direct and indirect speech in written language.The format of a direct and an indirect report may be identical except for quotation marks, and quotation marks may not be entirely reliable either (see example ( 11)).
Because of the thin line between a direct and an indirect quote, care is needed when applying labels as direct and indirect.If indeed deictic orientation were the only linguistic difference between a direct and an indirect quote, then there would be quite a number of reports that could not be considered to be direct or indirect since the deictic orientation does not appear on the surface.Thus, in the absence of quotation marks and deictic changes, it is impossible to put a label on the form of reported speech in example (13): (13) México arriba al nuevo milenio con sólidos cimientos de unidad, identidad y esperanza que ninguna adversidad puede destruir; así como con una democracia plena que permitirá edificar una vida digna, libre, justa y próspera para todos los mexicanos, aseguró el presidente Ernesto Zedillo en su mensaje de Año Nuevo a la Nación.(El Sol jan02prim2) 'Mexico arrives in the new millenium with solid foundations of unity, identity and hope that no adversity can destroy; as with a full democracy that will allow to build a dignified, free just and prosperous life for all Mexicans, assured president Ernesto Zedillo in his new year's speech to the nation.' One possible analysis would be that example ( 13) is an instance of direct speech in the sense that the quote is not introduced by que 'that' and the report verb comes after the quote.But it could also be an indirect report since no quotation marks are used.The above examples have shown, however, that quotation marks are not always reliable (example (11)), the position of the report verb is not fixed with an indirect report (different positions are shown in examples (1a) and (8a) ), and the indirect report does not need to be introduced by the conjunction que 'that' (example (8a)).Indeed, que 'that' may introduce a direct quote (example ( 9)).The methods section 4 will focus further on how different manifestations of reported speech were coded and how they are distributed.

Types of reported speech identified
Table 2 gives a summary of the forms of reported speech discussed above.The identification of the different kinds of reported speech was done by first considering deictic changes and, secondly, relying on punctuation both with the newspaper articles and the transcriptions of the sociolinguistic interviews.Six forms of reported speech were identified and, except for number 5 (Indirect speech with direct quote embedded), these forms were found in both corpora, although with different frequencies (for frequencies see table 3).
In table 2, 'V' stands for report verb; 'Q' stands for quoted utterance; "" stands for presence of quotation marks in text/transcription; 'a/o' stands for 'in any order', that is, the report verb V can precede, interrupt, or follow the quote Q.In the rightmost column of table 2, the numbers are given of the examples that illustrate the respective report strategy.The example for the freestanding strategy is given in ( 14), following the table.Also, an additional example of report strategy 6 [V+Q] is given after the table to complete the picture of the deictic nature of that structure.
The freestanding report strategy, a term borrowed from Cameron (1998), refers to a form of direct report that is not introduced by a report verb, as exemplified in ( 14) where the informant in answer to a question reports an entire dialogue between her and her son without once introducing the respective quoted speakers: Inf. -Sí, muy bonito.Muy bonito es; me pareció hermoso.Por cierto que aquí, se puede decir -ya lo he platicado en varias veces -José, el que acabas de ver, bueno, me echó de la casa.Otra persona.-¿Por qué, señora?Inf.-(a)"Pues sí, madre; tú nunca has salido, tú tienes que ver algo.Tu hija te necesita, te está diciendo que vayas".(b)"Mira, Pepe, que yo no puedo, que no hay dinero.Que cómo dejo la casa, que abuelita, que el niño, que ustedes, quién sabe que tantas cosas.¿Cómo le hago, Pepe?"(c)"Tienes que ir, madre".( d Examples of a report verb and quote in any order are shown in (8a) and ( 13).An example of an indirect report is (8a).An example without deictic change is (13); thus, we cannot say whether it is a direct or indirect quote.Example ( 15) is a report in the same format [V + Q], but it is a direct report as can be seen from the deictic perspective of creo 'I believe', where the words are interpreted from the point of view of the reported speaker.
(15) Creo, añadió el aspirante presidencial priísta, que lo que tenemos que hacer es presentar -y es lo que ellos deben hacer--propuestas alternas que superen las que nosotros hemos presentado en estos días.(El Sol Jan06prim5) 'I believe, added the candidate of the PRI for president, that what we need to do is present -and that is what they should do -altenative proposals that are better than the ones we have presented these days.'

Hypotheses
Direct and indirect speech are two strategies that have a function in common: they allow a speaker or writer to report words uttered at some other time.A question is, why two rather than one?This article examines the two hypotheses already discussed.The first hypothesis is that direct speech is an 'easier' strategy than indirect speech.Li (1986) suggests that it is easier to imitate (direct speech) than to paraphrase (indirect speech).Li (1986: 40) gives the following explanation: [...] indirect speech is a more complex communicative strategy than direct speech.Direct speech involves reproducing or mimicking the speech of the reported speaker, whereas indirect speech involves rephrasing or paraphrasing the speech of the reported speaker.Clearly, mimicking is a simpler undertaking than paraphrasing.Hence, it is not surprising that for reported speech the mimicking strategy occurs in all languages, whereas the paraphrasing strategy does not.
The problem that arises with such a claim is that it implies that direct and indirect speech are based on real dialogues and that if we report directly we really would report literally what was said.It is obvious that this cannot be true since we also use direct speech to anticipate dialogues (They will say "…..") or as a stylistic device in narratives, for example.Since we can report hypothetical utterances both directly and indirectly and since psychological research (Lehrer 1989) has indicated that human memory is not actually capable of recollecting an exact utterance.We remember the deep structure rather than the surface structure of an utterance and we must conclude, therefore, that both direct and indirect reports are paraphrases or constructed utterances of a deep structure we recollect.There is, thus, no difference between direct and indirect speech with regard to paraphrase.Unless one memorized an utterance on purpose.memory does not play a role except for recalling the content of an utterance which is the same for both direct and indirect speech.
If we, then, interpret Li's claim somewhat more liberally, namely imitating as pretending to be someone else even in the case that this person does not exist or is an animal, and interpret imitating as taking the point of reference of the person reported, we might expect this strategy to be easier in the sense that the reporter has only one point of reference.The reported utterance forms a clear break, deictically, with the ongoing communicative setting.Reporting indirectly, on the other hand, requires the reporter to bridge the two points of reference, his own (ongoing communicative setting) and the point of reference of the reported speaker (original communicative setting).This hypothesis will be tested by looking at the deictic orientation of the quotes: is the deictic orientation of the quote linguistically visible or not?That is, if we were to convert a direct report to an indirect report or an indirect report to a direct report is there any deictic element that would need to be changed?If so, the utterance is considered [+ deictic change] (whether direct or indirect), if not, the utterance is [-deictic change].Assuming, as explained above, that the speaker/writer has to 'reconstruct' the utterances s/he is reporting, it is expected that a speaker/writer will avoid deictic accommodations if possible.That is, the more the original communicative setting and the ongoing communicative setting differ deictically the more likely reporters will choose the direct reporting strategy.
If the hypothesis that there is a relation between [+ deictic change] and direct speech and [-deictic change] and indirect speech, is indeed supported, the question remains: When do people report directly and when do people report indirectly if no deictic change is needed?That is, if 'easiness' is not a motive, what is?To account for those cases, and for counterexamples, a second hypothesis is proposed: direct speech has a more involving character than indirect speech has.Chafe (1982) discusses detachment and involvement in written and spoken language.He finds that written language has a "detached quality", that is, the relation between writer and audience is one of detachment.Examples of linguistic devices of detachment he gives are the passive or impersonal constructions and relative clauses.Spoken language has more characteristics of involvement, that is, the relation between the speaker and the audience is one of involvement.Linguistic devices of involvement are, for example, emphatic particles ('just' or 'really') and direct speech (Chafe 1982: 45-48).Reporting directly, the speaker 'plays the role' of the reported speaker (by taking his/her perspective) and draws the attention of the interlocutor who needs to pay attention in order to interpret the deictic elements, such as 'yesterday' or 'my sister' correctly.This is similar to the relation between listening to X tell you directly an experience s/he has had (this direct strategy is what I refer to as "more involving") or have another person re-tell you the experience (this indirect strategy of paraphrasing is what I refer to as "less involving").Reporting directly, the reporter reconstructs a 'dialogue' and makes the interlocutor a direct experiencer, hearing or reading the dialogue from the perspective of the original communicative setting, whereas reporting indirectly, the interlocutor hears or reads the story from the perspective of the ongoing communicative setting.Or as formulated more generally by Goffman (1974: 503): A tale or anecdote, that is replaying, is not merely any reporting of a past event.In the fullest sense, it is such a statement couched from the personal perspective of an actual or potential participant who is located so that some temporal, dramatic development of the reported event proceeds from that point.A replaying will, therefore, incidentally be something that listeners can empathetically insert themselves into, vicariously reexperiencing what took place.
The involvement hypothesis will be tested by analyzing the following factors: 1) number of reported speakers: is there one reported speaker or are more people reported at the same time?; 2) the person of the reported speaker (first or third person); 3) individuatedness of the reported speaker, that is, can the reported speaker be linked to a specific individual (Maria, my parents, etc.) or not (the church, parents, etc.)?; and 4) historical vs. non-Historical Present tense value of the report verb.These factors and their relation to involvement are explained in further detail in the analysis.
After a description of the material and methods in section 4, the above hypotheses will be tested with quantitative analyses.

Material and methods
The data used in the analysis come from two corpora: a written and a spoken corpus.The written corpus consists of online newspaper articles from the Mexican newspaper 'El Sol de México'.The articles on the front-page from the first week of January 2000 were used (excluding newspaper genres such as editorials or opinions to keep the corpus homogenous).The spoken corpus consists of 14 thirty minute sociolinguistic interviews with both men and women from different age groups from Mexico City (Centro de Lingüística 1971).
Table 3 presents the number of cases encountered for each of the report strategies in the two genres respectively.In line with Chafe's analysis of detachment and involvement (1982), we find a difference in distribution of direct and indirect speech with regard to the two genres: direct speech (strategies 2, 3 and 4) occurs more frequently in conversations (75.2% vs. 25.6% in the newspapers) whereas indirect speech (strategy 1) occurs more frequently in the newspapers (45.1% vs. 23.6% in conversations).
These opposite tendencies may be due to various factors.It might be due to the different communicative goals that are pursued when speaking about one's personal life and work, and when writing about what is happening in the world.It might be due to the fact that we have more time when we write than when we speak (although newspaper writing may not be the best example of writing without time pressure, it still is a slower process than speaking (Chafe 1982: 36)).If indeed indirect speech is a more difficult strategy than direct speech, the fact that writing is a slower process, allowing for more reflection, than speaking might be a reason of a higher frequency of indirect reports in the newspapers.
Not only between the two genres but also within the corpus of conversations different frequencies of direct and indirect speech were found.The different articles in the written corpus however all showed similar distributions.The differences in the spoken corpus may be due to differences in topic of conversation, personality, relation between participants, or even the mood the informant was in that day; differences that are more likely to surface in spoken language than newspaper writing, which appears to be a fairly stable genre.For the moment, however, it is not expected that differences in distribution of direct and indirect speech imply that that their function differs.Rather it is expected, as is argued by Finegan and Biber (1994: 315): 1), "that the distributional patterns of these [linguistic] features across situations can be motivated functionally and 2) that such features function in comparable ways for all members of a speech community."Thus, direct and indirect speech, it is assumed, each have their own specific function and will be used accordingly by speakers and writers of specific genres.Differences in the distribution between speakers/writers or intraspeaker variation can be explained by the communicative goal they pursue and thus the need for language speakers to draw upon one or the other strategy (in this case, direct or indirect reported speech).
In table 3, three forms of direct report strategies are given: 2 Direct speech [V+"Q"]; 3 Freestanding ["Q"]; 4 Direct speech introduced by que [V+que+"Q"].Because of their same deictic orientation they will be taken as one category in the quantitative analysis, where possible.If, however, the distribution of one of these strategies deviates from the others this will be mentioned.Because of their low frequency strategies 5 and 6 will not be analyzed quantitatively in the next section.

Report verb
Several different report verbs were found accompanying the direct and indirect quotes.Table 4 presents the most frequently report verbs in the corpora investigated.Some of the report verbs were found to accompany both direct and indirect quotes; others, however, were found only with one or the other.Contar 'to tell', garantizar 'to garanty' and recomendar 'to recommend', for example, were only found with indirect reports, agregar 'to add' and argumentar 'to argument', on the other hand, were only found with direct reports.It may turn out that some report verbs are used only with the direct or indirect strategy.

Forms of reported speech not included in the quantification
A form of report that has not been included in the quantification is a form in which the 'quote' is the direct object of the report verb without being subordinated to it.It seems to be an abbreviated form of indirect speech, as exemplified in ( 16): (16) Así lo resolvimos y le dijimos al chofer que si él nos podría llevar.Le dijimos dónde y todo.Él, muy satisfecho, contestó que sí, que cómo no, que nos llevaba con toda facilidad, y que... Nos arreglamos en el precio, y ya, nos fuimos muy bien.(Centro de Lingüística 1971:124) 'That's how we solved it and we said to the driver that if he could take us.We told him where and everything.He, very satisfied, replied that yes, that why not, that he would take us without any problem, and that … We arranged the price, and we went already.' Similar to the example above is (17).In ( 17), however, there is a reported utterance and the reported speakers can also be identified.It is not clear, however, to which of the categories in table 3 they would belong.( 17) "Ustedes son la esperanza del mundo, la esperanza de la humanidad", aseguró el papa ante los infantes, quienes le respondieron con "vivas" y con coros de "Te queremos mucho".(El Sol Jan03prim2) ' "You are the hope of the world, the hope of humanity", assured the pope to the children who replied to him with "vivas" and choruses of "We love you very much".' The main reason to exclude examples of the type of ( 17) is that the reported quotes follow a preposition and appear to behave as nouns rather than actual utterances.This makes them different from the other report strategies and that there is only one example.The passage is interesting, though, because of the first report vivas (literally 'live', an exclamation of praise, reported in the plural).It is a rather unusual quote.It seems to be a direct report because of the quotation marks, but a direct quote would have had the form viva and not the plural vivas.Because of the different nature of these particular reports, I have excluded them from the quantitative analysis.
Examples ( 16) and ( 17) have thus been excluded from the quantitative analysis at this time.These examples show that there are many ways of reporting.Even ( 16), ('we told him where and everything'), is a way of paraphrasing what was said.

Analysis
In what follows, the data will be analyzed both quantitatively and qualitatively.The quantitative analyses are performed to examine the hypotheses postulated in 3.4.The qualitative analyses are expected to account for data in their specific context.In 5.1, the hypothesis that direct speech is a less complex strategy than indirect speech is tested; the hypothesis that direct speech is a more involving strategy than indirect speech is tested in 5.2.

Reported speech and deictic change
The first factor investigated is the deictic orientation of the quoted utterance.That is, is the deictic orientation of the ongoing communicative setting different from the deictic orientation of the original communicative setting with regard to the utterance?As illustrated in fragments ( 2) and ( 4), the deictic orientation of the quoted utterance might be the same as that of the original utterance regardless of whether the report is direct or indirect.
It is expected that if the speaker/writer reports directly, the quoted utterance contains deictic elements (including tense) that would need to be adapted to the ongoing communicative setting if they were reported indirectly, as for example in (1b), repeated here in (18) for convenience: (18) Y que estábamos rezando por la noche los... rosarios, y cuando salíamos, que... habíamos terminado de los rosarios, me dice en el corredor... me puso la mano en el hombro, (b) y me dice: "Licha, está tu mamá lindísima, gloriosa, en el cielo."(Centro de Lingüística 1971:143) 'Imagine that my mother died three days before her son.And at night we were telling our... beads, and when we left, that... we had stopped telling the beads, he says to me in the corridor... he put his hand on my shoulder, and he says to me: "Licha, your blessed mother, is in heaven.' In order to report (18) indirectly the reporter would have to leave out the vocative Licha, and change the possessive pronoun tu 'your' to mi 'my' and possibly the verb está 'is' to estaba 'was'.On the other hand, the hypothesis predicts that if the speaker/writer reports indirectly, the quoted utterance does not contain deictic elements (including tense) that would need to be adapted to the ongoing communicative setting (see (5) for an example).Table 5a presents the results regarding the relation between deictic change and report strategy.5a shows that there is indeed a high correlation between direct speech and the need for a deictic change if the quote were to be rendered indirectly (75.3% deictic change needed vs. 30.8%no deictic change needed).The opposite is true of indirect report and no deictic change if it were to be rendered directly (69.2% no deictic change needed vs. 24.6%deictic change needed).
In table 5a the two genres have been taken together and for direct reports the categories [V+"Q"] & [V+que+"Q"] and the freestanding strategy ["Q"] have also been taken together.There might be some genre differences, however, between the conversations and newspaper articles regarding the Freestanding strategy.Table 5b gives more detail.The Freestanding strategy in conversations is used more frequently if deictic changes are necessary if reported indirectly (23.9% vs. 7.7%), which supports the hypothesis; in the newspaper articles, there is a tendency to use the Freestanding form although no deictic changes would need to be made if reported indirectly (albeit a very weak tendency: 5.9% deictic change needed vs. 8% no deictic change needed).
A separate analysis of [V+que+"Q"], also indicates that this strategy has a very similar distribution as indirect speech introduced by que [V + que + Q] has regarding the factor of deictic change.Although there are only 15 cases of the format [V + que + "Q"] included in the quantification, 12 of those (73.3%)contained a quote that would not have changed if reported indirectly (that is, if the punctuation had been omitted).Nearly one third of the direct reports that indicates no deictic changes, thus, comes from this particular strategy introduced by que It can be concluded that, with the exception of the strategy [V + que + "Q"], direct report strategies tend to contain deictic elements that would have needed to be accommodated if reported indirectly.That is, by using a direct report the speaker has 'avoided' this deictic accommodation.This supports the hypothesis that direct speech is an easier, more economic strategy than indirect speech or, formulated differently, that if the original communicative setting is deictically different, direct speech is the preferred linguistic strategy.

Direct speech is a more involving strategy than indirect speech
With regard to the second hypothesis that direct speech is a more involving strategy than indirect speech, four factors are analyzed: 1) number of the subject of the report verb; 2) person of the subject of the report verb; 3) individuatedness of the subject of the report verb, and 4) Present tense value of those report verbs that are in the Present tense.

Number
The first factor investigated to test the hypothesis that direct speech is a more involving strategy than indirect speech, is the number of the subject of the report verb.In other words, can the quoted utterance be attributed to one or more original speakers?
According to Coulmas (1986), it would only be possible to quote a group of people directly if all spoke in chorus and still then it is problematic because a single reporter cannot play the role of more than one reported speaker at a time 4 .Reporting more than one person, therefore, implies a de re interpretation.Based on the de dicto / de re analysis and its implications regarding involvement as discussed in section 2.1 one would, thus, expect speakers to use more indirect speech when more than one person is reported and more direct speech if one person is reported.If indeed direct speech is a more involving strategy, one would expect more singular original speakers with direct speech and more plural original speakers with indirect speech.Reporting more than one person at a time directly, means that the dramatic force of direct speech gets lost since one cannot impersonate more than one person at a time.There is, thus, no reason to use the direct strategy because the reporter cannot profit from its involving character.
The following passage shows a speaker reporting more than one person with an indirect report and a single person with a direct report.Example (19) shows how the journalist reports basically the same information no se registró ningún problema 'no problems were registered' and "Pasamos la medianoche en Washington sin ningún problema del Y2K" '"We spent midnight in Washington without any Y2K related problem"' first indirectly with a plural reported speaker (a) and then directly with a singular reported speaker (b).
The classification of the number of the original speakers is based on the conjugation of the report verb.The freestanding strategy has therefore been excluded from this analysis since it does not have a report verb.The results of this analysis are presented in table 6.  6 shows that there is indeed a statistically significant relation between a singular subject and direct speech (60.0%singular vs. 31.9%plural subjects) and between plural subject and indirect speech (68.1% plural vs. 41.0%singular subjects).However, there are counterexamples.For now, I leave aside the singular subjects because the next two sections will give more information regarding these.A counterexample, (20), is given below regarding the report of a plural subject with direct speech.
(20) Por fin, ya con todo esto, el viaje ya se hizo muy largo, y pensábamos llegar a Métepec... pues, a las cuatro de la tarde o las cinco, todavía con luz.Pero llegando a un lugar donde teníamos que trasbordarnos... a otro camión, ya era de noche.Nos bajábamos del coche aquel para ir a tomar mejor un camión, en vista de las dificultades que llevábamos, y al bajarse, a una de mis sobrinas, a Angela, se le abrió la petaca, porque no iba cerrada.Una petaca donde llevaba sus útiles de tocador, se le abrió, y regó todo lo que llevaba.Ahí llevaba unas joyitas, para adornarse, y... y a esas horas, en la oscuridad, ahí nos tienen todos a gatas, buscando todo lo que había tirado.Y otro motivo de... de momento de contrariedad, pero después de risa, porque en la oscuridad todos decíamos: "Qué daríamos ahorita por un fósforo, para poder encontrar todo lo que hace falta!" (Centro de Lingüística 1971:125) 'And finally, with everything, the trip had taken a long time and we thought we would arrive at Métepec… well at four p.m. or five p.m., it was still light.But arriving at a place where we had to change to another bus, and it was already night.We got out of that car to better take a bus, considering the problems we had had, and while getting out one of my nieces, Angela, her suitcase fell open, because it was not closed.A suitcase in which she had stuff from her dressing table, it fell open and she spread everything.She had some jewelry, to adorn herself, and… and that late, in the dark, there you have us all on all fours, looking for everything that had fallen.And another motive of … of a moment of contrariety, but after laughing, because in the darkness we all said: "What would we give at this moment for a match to be able to find everything that is missing!" ' In (20), it is unlikely that the speaker and her nieces said the quoted utterance in chorus with a verb in the first person plural daríamos 'we would give'.But the reporter, by reporting directly, presents the event in a context of consensus to her interlocutor.The reporter recreates the event as an amusing anecdote (después de risa 'after laughing') and she gives details so the interlocutor can recreate the event.The interlocutor has, thus, been introduced to the original communicative setting, and the whole setup of the story is such as to involve the interlocutor.A direct report, although it has a plural subject, follows naturally.It is the climax of the story in which the reporter and her nieces are together in the same situation and are, therefore, reported as if they were speaking in chorus.

Person
The second factor analyzed is the person of the subject of the report verb with regard to the report strategy.A reporter can report him/herself ('I said') or some other person ('Mary said').For this analysis only the spoken corpus is used since in the newspaper articles, as one may expect, only third persons were reported.
If the hypothesis that direct speech is a more involving strategy than indirect speech is correct, it is expected that we should find more direct reports when the reporter is the same person as the reported speaker, that is, the subject of the report verb is the first person, and more indirect strategies if the reporter is someone else.Although it is not necessarily true that the interlocutor feels more involved when the reporter is speaking about him/herself (although, having the protagonist tell the story right in front of you might be very involving), it is fairly likely that the reporter will feel involved when talking about him/herself and will report his/her own words directly instead of indirectly.Table 7 presents the results of the analysis of the relation between person and report strategy.For examples regarding the expected distribution, see (20), where the reporter reports herself directly ('I said') and (1b) where the reporter reports a third person ('he said') Since the Freestanding strategy does not have a report verb this strategy cannot be taken into account.Very few second persons were quoted in the articles analyzed and were therefore excluded from table 7.  7 shows that there is a correlation between the direct speech strategy and the first person (84.8% first persons vs. 62.9% third persons) and between the indirect speech strategy and the third person (37.1% third persons vs. 15.2%first persons).As expected, there are very few cases where the reporter reports his/her own words indirectly; however, within the third person reports there also is a higher percentage of direct reports.The next analysis, regarding individuated vs. non-individuated subjects, sheds more light on the relation between third persons and the report strategy used.

Individuation of the subject
The individuatedness of the reported speaker plays an important role in the distribution of direct and indirect speech.Individuated subjects are those subjects that could be 'matched' with one or more individuals (proper names, the president, my sister, her parents).By the same token, non-individuated subjects are those that could not be matched with one or more individuals (people, everyone, it is said).An example of each is given in ( 21), where (21a) is an indirect report with a non-individuated subject (me dicen 'they tell me' no clear referent for 'they' is found in the context) and ( 21b) is a direct report with an individuated subject (me dijo 'she said to me' where she refers to the niece of the reporter).( 21 If the hypothesis is correct that direct speech is a more involving strategy, it is expected that there is a correlation between direct speech and an individuated reported speaker.By contrast, if it is correct that indirect speech is less involving, a correlation between indirect speech and a non individuated reported speaker is expected.With regard to the freestanding strategy, it resulted from a separate count that of a total of 75 cases, in 77.3% the reported speaker is individuated, and in 22.7 % of the cases it is not.The conclusion that can be drawn is that the relation between the freestanding strategy and the individuation of the subject supports the hypothesis that direct report strategies are more likely than indirect strategies when the speaker reported is individuated.

Present tense value of the report verb
The last factor investigated to test the hypothesis that direct speech is a more involving strategy than indirect speech is the Present tense value of the report verb.As is well known, the Present tense is not only used to express what happens right now, but it can also be used instead of the Preterite, in which case it refers to the past.This use of the Present is called 'Historical Present' used to "refer to events which began and ended at some time previous to the moment at which the narrative itself is told" (Wolfson 1982).In this definition we can see there is a link between Historical Present and direct speech; direct speech, namely, often also refers to an event (or dialogue) "which began and ended at some time previous to the moment at which the narrative itself is told."Chafe (1994: 208) pointed out that "Speakers in the displaced mode [consciousness focused on experiences that are derived from another, earlier consciousness, not the immediate environment] may pretend to be representing experiences that are closer to those of an extroverted consciousness [consciousness focused on the immediate environment] in either or both of two ways.One device of this sort is the Historical Present; the other is direct speech."Thus, both direct speech and the Historical Present have in common that the speaker remembers or imagines what might be present in the extroverted consciousness of the reported speaker.A similar evaluation seems to underlie Silva-Corvalán's (1983) attribution of an internal evaluation function, rather than a rhetorical function to the Historical Present: "By using the present to describe events which occurred in the past, the speaker presents them as if they were occurring in front of his eyes.This creates the effect of immediacy and makes the narrative more vivid and dramatic."(Silva-Corvalán 1983: 774-775).
Two examples of reports introduced by a report verb in the Historical Present are given in ( 22), also other verbs in this fragment can be found that are in the Historical Present and are underlined.In ( 22), the speaker (Carmelita) tells an anecdote of how her husband (Juanito) tried to get her attention the first time.
( In ( 23) the present tense is not considered to be a Historical Present, because in this case, although the words that this newly discovered temple is fabulous were uttered at some moment in the past they have not lost their actuality as the utterances in (22a) and (22b).
If indeed direct speech is a more involving strategy than indirect speech, a relation is expected between the Historical Present and direct speech, and between the non-Historical Present and indirect speech.The use of the Historical Present facilitates telling about the past in such a way that the interlocutor can recreate the story and feel as if s/he is a witness.The non-Historical Present does not have such dramatic force, either it describes something that goes on at the moment of speaking or something of 'universal value', but the interlocutor does not need to imagine a different communicative setting as it would be the case when a history teacher tells about what happened when the Romans invaded Jerusalem, for example.8 shows that the Historical Present was only found in the conversations.Further, all report verbs in the Historical Present introduced a direct report.In fact, in the written corpus only six report verbs in the present tense were found, all introducing an indirect quote.The corpus of conversations, the only one with variation, also contains 12 cases of direct reports introduced by a report verb in a non-Historical Present.One such example is given in (24): (24) Enc.-¡He tenido una mala suerte con mi tía Elena!Fíjate que yo la he de haber conocido cuando era yo muy chica, pero siempre tuve yo intención ahora de irla a ver.Pero... pues, de eso que se te va pasando el tiempo... no sé qué pasa.It may be concluded from table 8 that there is indeed a high correlation between direct speech and the Historical Present (100% Historical Presents introducing direct speech vs. 26.7%non-Historical Presents introducing direct speech) and between non-Historical Present and indirect speech (73.3% non-Historical Presents introducing indirect speech vs. 0.0% Historical Presents introducing indirect speech).
One question that needs more attention is the relation between both hypotheses.It might be the case that original and reported dialogues that have few deictic elements in common are more involving than original and reported dialogues that have (nearly) identical communicative settings and, thus, have all deictic elements in common.The interlocutor has to pay more attention if the original and reported dialogues have few deictic elements in common in order to correctly interpret the deictic elements if the utterances are reported directly.However, in the present study no distinction was made between the different kinds of deictic elements that can change; that is, no distinction was made between tense and adverbs, or pronouns and demonstratives.One could argue that sometimes a pronominal change or a change in tense is, actually, more complex if it breaks the tense harmony between the current and original communicative setting, as in example (25b) as compared with (25a).
(25a) Y de que me dijo Carmela que se iba con las madres de las yeguas finas... (Centro de Lingüística 1971: 134) And about that Carmela told me that she would be going with the mothers of the 'yeguas finas'… [nickname for a convent] In (25a), the subject of the report verb is the same as the subject of the main verb in the quote and both verbs are in the past tense.A direct quote may have looked like (25b): (25b) Y de que me dijo Carmela: Me voy con las madres de las yeguas finas… And about that Carmela said to me: I will be going with the mothers of the 'yeguas finas '… In (25b) where there is a shift from third person and a past tense in Carmela dijo 'Carmela said' to the first person and a Present tense me voy 'I will be going', a shift that breaks the harmony past-past and third person-third person.It might be that there are different degrees of complexity or even that certain deictic changes prevent complication in the sense that they do not break a harmony in tense or personal pronouns.

Conclusion
In this article two genres were investigated, conversations and newspaper articles which each have a different communicative goal.In the sociolinguistic interview the conversations aim to be like 'natural conversations' and we would expect them to have a more 'involving' nature.Newspapers on the other hand are foremost informative in nature.The study examined the effect of genre on the distribution of direct and indirect reporting strategies.The article finds that the reporting strategies used reflect these different communicative goals, in that direct speech is more frequent in the conversations whereas indirect speech is more frequent in the newspapers.
Two hypotheses were tested: 1) the hypothesis suggested by Li (1986) that direct speech is an 'easier' strategy than indirect speech and, 2) that direct speech is a more involving strategy than indirect speech (suggested by Chafe 1982;Tannen 1989, among others).I hope to have demonstrated that there is indeed statistical support for these hypotheses.'Easiness' of the direct report strategy has been supported by the analysis presented in 5.1, where a correlation was found between direct speech and [+ deictic change] (that is, there would have been deictic change if it were to be reported indirectly) and indirect speech and [-deictic change] (that is, there would not have been deictic change if it were to be reported directly).Some genre differences were found as well with regard to the freestanding category of direct speech being used more in newspapers when no deictic change is necessary.The second hypothesis, that direct speech is a more involving strategy than indirect speech, has been supported by the fact that there is a correlation between direct speech and singular, first person and individuated subjects of report verbs, and the co-occurrence of the Historical Present and direct speech.
While we may not know if easiness is in fact the relevant cause, what we can observe more descriptively is that if the original and ongoing communicative setting differ deictically, that language users tend to use direct speech as the preferred linguistic strategy.Or alternatively, when reporting directly, one tries to involve the interlocutor/audience and invoking the original communicative setting in all its deictic detail is the best way to do so.

(
19) (a) Funcionarios militares rusos y estadounidenses trabajando lado a lado en un centro de comando conjunto en la Base Peterson de la Fuerza Aérea en Colorado Springs, en el estado de Colorado, manifestaron que no se registró ningún problema.(b) "Pasamos la medianoche en Washington sin ningún problema del Y2K", aseguró el mayor de la Fuerza Aérea Thomas Goslin, director de operaciones del Comando de Estados Unidos en Colorado.(El Sol Jan02prim2) (a) 'Russian and North-American military officials who work next to each other in a command center in the Peterson Base of the Airforce in Colorado Springs, Colorado, manifested that no problems were registered, (b) "We spent midnight in Washington without any Y2K related problem", assured the Air Force major Thomas Gosling, director of operations command of the United States in Colorado.'

Table 2 .
Overview of the different types of reported speech analyzed ) "Mira, pero fíjate que el pasaporte, y todo lo que tengo que hacer; yo no tengo ganas".Porque quedé sin ganas de hacer nada, señorita... -Carmen se llama, ¿verdad?(Centro de Lingüística 1971:82) [Talking about New Orleans] 'Inf.-Yes, very nice.It is very nice; I found it beautiful.Certainly that here, one can say -I have already told it various times -José, the one you just saw, well, he threw me out of the house.Other person.-Why ?(a) "But yes, mother; you have never gone away, you have to see something.Your daughter needs you and is telling you to go".(b) "Look, Pepe, but I can't, there is no money.And how do I leave the house, what about grandma, what about the child, what about you, there are so many things.How do I do it, Pepe?" (c) "You have to go, mother".(d) "But look, what about the passport, and all I have to do; I don't feel like it".Because I didn't feel like it, miss … -Your name is Carmen, isn't it?'

Table 3 .
Frequency of types of reported speech

Table 5a :
Frequency of types of reported speech according to whether any deictic changes are needed if converted to the other report strategy

Table 5b :
Frequency of types of reported speech according to whether any deictic changes are needed if converted to the other report strategy; divided by genre.

Table 6
Types of reported speech by number of the subject of the report verb

Table 7 :
Types of reported speech by person of the subject of the report verb in the

Table 8 :
Types of reported speech by individuatedness of third person subjectThe distribution shown in table 8 also supports the hypothesis.If the reported speaker(s) can be identified, their words tend to be reported directly (54.7% individuated vs. 34.1% non-individuated reported speakers).If the reported speaker(s) cannot be identified, their words tend to be reported indirectly (65.9% non-individuated vs. 45.3%individuated reported speakers).
22) Y un día me llevó un libro, que todavía tengo ahí, que se llama La hija del director del circo, que tiene nada más... "Lolita, ¿qué, me permite usted regalarle a Carmelita este libro?"(a)Y le dice ella: "Pues, siendo de usted, Juanito, no creo que le regale usted algún libro que no le... debiera leer.Sí, cómo no".Y yo que hago así, y que veo adentro una carta.Me abre Lolita el libro que... Y el otro día voy con mi padre director, y antes de leerla yo, la lee él.Both (22a) and (22b) are reports of words uttered at some point prior to the ongoing communicative setting.The present tense is, thus, used as a Historical Present.An example of a report introduced by a report verb in the present tense that is not a Historical Present is given in (23):(23) Tú sabes que hace poco han descubierto un nuevo palacio que se llama... ¿cómo?...
(b) Dice: "Pues esto es una decl... se declara, es una declaración de amor; mira".Ahí la tengo guardada todavía la carta.(CentrodeLingüística 1971: 133)'And one day he brought me a book, that I still have, that is called the daughter of the circus director, that has nothing… "Lolita, do you permit me to give this book to Carmelita?"(a) And she says to him: "Well, since it is you, Juanito, I don't think you would give her a book that she….she should not read.Yes, why not".And I, that I do like this, and that I see a letter inside.Lolita opens the book that… And the other day I meet with the father who directs me and before reading it myself he reads it.(b) He says: "Well this is a decl… it is declared, it is a declaration of love; look".I still have the letter.'

Table 8 :
Types of reported speech by Present tense value of the report verb .Inf.-Sí; no sé qué pasa, que dice uno: "Sí... Voy a verla, y voy a verla", y nunca va uno.(Centro de Lingüística 1971:130) 'Enc.-Ihave had bad luck with my aunt Elena!.Imagine that I must have known her when I was very young, but always I had the intention to now go visit her.But… well, like that time passes by on you… I don't know what happens.Inf.-Yes;I don't know what happens, that one says: "Yes… I/m going to see her, and I'm going to see her", and one never goes.'