1. Introduction
The authenticity of reported speech has been contested since Volosinov (1973 [1929]) , who stressed the altering nature of original speech with the one reported—a view that has been developed by successor researchers, who defined it as “constructed dialogue” or as “misleading term” (Tannen, 1989, 2007) , “pseudo-quote” (certain constructions: Dubois, 1989: pp. 347-348 ), “untearable from original habitat” (Sternberg, 1982) . Subsequently, Mayes (1990) in a study of conversational narratives and ordinary interaction (320 cases in total) shows that at least 50% of direct quotes in informal spoken discourse in her corpus were doubtful on their originality, “some of them unquestionably, invented” (p. 331). When a speaker uses reported speech to portray the words of another, he or she is the animator of the utterance, but no longer the author or principal (Holt, 2007: p. 48) , because there is no guarantee about accuracy of transmitting exact way something was uttered, unless it was recorded.
Thereby, it’s a true fact that a past quote is unrepeatable originally by a person, as is quite impossible to cut and recontextualize the authentic piece of discourse & reproduce it in strong accuracy rate in other contexts. Little shifts may happen unintentionally, though efforts are maximal to remain faithful, or the statement may be twisted when RS contains “creative details” which fit with speakers’ convenience in order to trigger desired outcome and to show convenient nuances, but its “subtract” can remain relatively true (partly based on real occurrence). However, the speech also may be completely restructured (though situation may be real) or totally imagined purposely. A statement and other connotative nuances or assessments may be falsely attributed to someone and convince the hearer about its truthfulness. The attempting to deceit can be detected through (non)verbal signs with various degree of effectiveness, though is not proved total assurance for effective/clear formula to detect lies (Vrij et al., 2019; Vrij, 2018; Vrij, 2008; Houston et al. 2012; Meibauer, 2018; Newman et al., 2003; Meyer, 2010) . When the overall performance of speaker doesn’t fit with our general conviction about the reported speaker (his/her behaviors, linguistic expressions in our episodic memory), we end asking ourselves: did s/he really say that?
In this research we will identify pragma-linguistic mechanisms that signal the (non)truthfulness of a reported speech, which is once delivered by respondents and then, is confirmed by them if it was true or not. Based on declared (non)veracity, we’ll collect crucial clues involved on (un)true reports and we’ll present the most prominent ones in this paper.
Despite the large interest and contribution of scholars in topics of lying and reported speech, there haven’t been studies around the potential cues that should be monitored in order to gain insight about its reliability. Thus, the reason that prompted this research is to yield incentives upon which we can rely on, based on empirical methods, and see the interface between these psychic and linguistic concepts. Within this scarcity, McGlone & Baryshevtsev (2018) in their study Lying and Quotation have stressed the fact that quotations can be altered, certain phrases can be removed, or only a portion of it can be reproduced.
According to Vrij (2018: p. 308) , “truth-tellers include more speech in its original form (quotes) than liars”. However, we’ll discuss how both: liars and truth-tellers formulate pieces of past discourse when they want to convince the hearer about its veracity.
After giving definitions/explanations of what is a lie in communication, we’ll present parts of conversations and certain devices used when falsely or sincerely past declarations of others are uttered. After that, we’ll briefly explain the main detected parts that “shield” RS in order to incline or prevent emotional reactions. Next section outlines usage of main characteristics of utterances at conversations and compares them according to the veracity of RS.
In the last part, our primary objective is to describe statistically and shed light on prominent/emphasized pragma-linguistic constructions and modal words that function as resources to perform lies or to report sincerely the utterances of others.
2. (Dis)Honesty in Communication
It isn’t common in normal relationships to directly call someone a liar, or to tell him/her that s/he’s lying (cf. Ekman, 1992 ), unless it is a public confrontation between two opposite sides and the debate escalates in offenses. Even when the hearer clearly notices it, though s/he may challenge the speaker by demanding additional information or evidence that exists, it is considered “impolite and rude expression” to immediately give such epithets. So, the hearer may keep to himself the judgement of veracity, especially on polite/formal interaction.
One of the maxims in conversation that Grice (1989 [1975]) proposed, is the quality of utterance, which suggests that one does not have to give information that is false or that is not supported by adequate evidence, and on the other side, invite audience to trust that his intention is to say the truth. By violating this maxim, speakers may lie about someone’s statements or twist them deliberately to trigger certain reactions that strategically orient the hearer towards speaker’s intentions. The purpose, as we will see, may be to cause nervousness, frustration or their intent may be good one: to evade conflicts between persons, to calm down the hearer etc.
But, how/when do we start to produce untrue assertions? Does it require extraordinary skills? There have been extensive studies about lying on early age (Talwar & Lee, 2002; Lewis, 1993; Talwar et al., 2004) . “Before age 2; 6, children do not seem to be capable of lying” (Meibauer, 2018: p. 369) , thus they start quite early-as a preschooler (Peskin, 1992; Talwar & Lee, 2002; Talwar et al., 2007; Lee et al., 1997) . Some behaviors such as faking a cry/laugh, distracting adults may start on the 6th month (Meyer, 2010: p. 31) , while moral understanding of lie increases with age (Talwar et al., 2004: p. 429) . As an ability, is mustn’t be too hard to grasp, moreover, it may be considered as innate to human/animal nature ( Meyer, 2010 ; see also Darwin, 2009: pp. 133, 379-381 [1890] ), because of the creativity/imagination as an early characteristic of a human being, and also the desire to benefit preferred things/foods urges the child to confess untruthful claims, though they are easier detectable. Moreover, based on divergence of personalities, it doesn’t develop with age as a skill, because persons don’t lie in the same, neither similar level or manner, as e.g. the language is acquired or the body is developed similarly at normal individuals. Nevertheless, it is cultivated by becoming habit to a person, through scrutinizing/reinforcing strategies that work better and removing those which don’t function at certain individuals or not at all. But, persons do not necessarily improve them in the same rhythm, e.g. experienced liars or self-deceptive persons may be more confident on their skills, without exhibiting obvious/apparent cues. Thus, general assertions/assessments about development/strategies of lying at a person do not seem very convincing, nor reliable in real life (see Vrij et al., 2019 ), because the amount of lying isn’t distributed equally at people: that’s why we don’t give the epithet of a “liar” or “sincere person” to everyone on our circle, but we choose carefully who we trust on and call him/her “honest”, though everyone lies and everybody has little or enough evidence that has been sincere many times (even those who are known as deceivers), and we’re aware about that.
There are various possibilities to interpret the authenticity of a past statement (how much it is loyal with the exact version: in lexical, syntactic, pragmatic level) e.g., by trying to confirm through the reported speaker1 (there is no guarantee if s/he will be sincere about it), detecting (para)linguistic cues or other gestures. However, the reliance on the cues detected of course, is not absolute source to assess the truth-value of speech2, as many cues can lead us away from apprehending the truth. Some of them warn the speaker to make additional questions, ask for explanations, and thus, to draw a conclusion to him/her whether to believe or not.
What Is Lying?
Charles Darwin (2009 [1890]) in his book The Expression of the Emotions in Man and Animals, has emphasized the relation between intentions and movements/gestures which “are really expressive of certain states of the mind” (p. 14). Freud (1959 [1905]) shortly described the role of gestures on “speaking the truth”: “If his lips are silent, he chatters with his finger-tips” (p. 94). Thus, due to stress, nervousness of being caught, speaker may twist/contradict his own story, or start moving on strange manners, thereby speaking through his/her body. But how it is detected? Are there obvious cues that warn us about that?
Ekman and Friesen (1969) were the first who discussed more scientifically and influentially the nature of deception in interaction and its primary resources. They distinguished: alter-deception, “where S conceals information from the other interactant”; and self-deception, where “ego is the object of his own deception, concealing information from himself” (p. 89). They also classified deceptive clues and leakage ones, which many times are uncontrollable, cannot conceal totally the truth, because are manifested non-verbally. “Accuracy of lie detection includes both whether the detectors recognized that deception was occurring (deception accuracy) and whether they identified the concealed information (leakage accuracy)” (Zuckerman et al., 1981: p. 23) . Mastering these skills reveals inclination on filtering the proper information to reach higher chance on detecting the lie.
Williams (2002: p. 96) defines lying as “an assertion, the content of which the speaker believes to be false, which is made with the intention to deceive the hearer with respect to that content.” Carson (2006: p. 298) sees it as a falsehood that the speaker insincerely warrants to be true, while DePaulo et al. (2003: p. 74) as a “deliberate attempt to mislead others”, though this effort isn’t expressed openly. According to Simpson (1992) without this disguise, the lie couldn’t succeed, and without the intention that there be this disguise there would be no lie.
Lying in some contexts of everyday life and professional settings may become necessity, especially when a person is faced with difficult situations, and in such cases “another reality” would serve him/her some personal/group purposes or save him/her from a punishment and thus s/he presents it as true, knowing/believing secretly that it is false. E.g. even though “leaders rarely lie about verifiable facts” (Spicer & Spicer, 2018: p. 11) in political discourse, Trump was caught repeatedly “lying” even egregiously (see Spicer & Spicer, 2018 ). Also, after paying special attention, Bill Clinton’s words and facial expressions show him very suspicious after denying his affair with Monica Lewinsky (Meyer, 2010) .
Taking into account subjects to whom lies serve, there are: antisocial (mendacious) lying—only the speaker benefits from his lie and prosocial lying—with benevolent intentions: “gray lies, blue lies” (Hornung, 2016; Fu et al., 2008; Bryant, 2008) These lies for someone’s benefit or interest are antisocial, known as “malicious” and “self-serving” (Bryant, 2008: pp. 33-34) . Prosocial lies or white lies are performed with unharmful intentions, to “preserve feelings” and “save the positive face” (Fraser, 1993; Levine and Schweitzer, 2014; Borsellino, 2013; Hornung, 2016) . They can be altruistic (only the hearer profits from the speaker’s lie) or a group of persons (see Meibauer, 2018: p. 371 ).
A person decides to apply the lie, also being aware for her necessity to accomplish her goal, which is more important than to feel free from the potential label as a “liar” and the possibility to get caught, so s/he endures the stress and insecurity, for something “worth fighting for”. And to meet these demands of an individual, in order to form closer relation with someone or other reasons, when the truth is inconvenient or traumatic in many uncommon situations, people decide to lie. For instance, “the mother asserts a proposition, which she does not believe, yet wants her child to believe it” (Fraser, 1993: p. 151) . In such cases, “the family peace is threatened and she is afraid of the possible impact that could have on her child” (Hornung, 2016: p. 239) .
If s/he confesses her mistake of not telling the truth and the repentance, the hearer would still know that she lied, so if the discussion takes too long, there is more likely to display evidence on insincerity. E.g. after being asked in the court, the prosecutor knows how to dig deep on the issue to prevent the person from lying.
As a phenomenon, lying doesn’t only cause divorces, injures trust/feelings, end of friendships and many other conflicts, and even deaths (Meyer, 2010) , but it also causes unjust convictions by being causer of declaring the innocent as guilty (see Garrett, 2011 ), the faithful as traitor etc. Even though some “false assertions” are considered as little momentary lies, they may have serious consequences as they foster person to believe a false version of reality.
The declarations of others are one of constructions that often get twisted, changed or misleadingly modified in totally different way. We will present some concrete extracts, in order to examine the traceability of (un)real reported speech.
3. Data and Methods
31 conversations of (un)true reportings about sayings of a common friend (known by both: hearer and speaker) that they heard in real life, and those which were purposely created by them, were recorded in Prishtina (Kosovo). They are conducted by 45 respondents aged 23 - 28—where 68% of them were both: once reporters, then hearers/interlocutors, while 14 were only hearers. 14 of reports were true, 17 false. They were encouraged to be as persuasive as possible in surrounding elements of RS in both cases while delivering their reportings—not to give an overt signal that would obviously show truthfulness. In addition, they were instructed to utilize pragma-linguistic mechanisms they occasionally use when they lie or speak the truth in real life. To analyze degree of veracity, we urged them to tell something to their friend/colleague, and then confirm if it was a true reported statement or deception. Also, the interlocutor gave her/his degree of persuasion (in percentage), to see if s/he can accurately value degree of veracity. Their first respond was accounted as definitive, and the accuracy on detecting real truth-value was measured by degree of matching with the veracity of utterance of the respondents (confirmed by them alone) that produced the reported speech.
4. Cues of Deception and Sincerity in Reported Speech
In this section we will deal with three concrete (un)true utterances of reported speech which are analyzed carefully to detect the properties of a false report and a sincere one. Below are examined only a few conversations to demonstrate the prototypical pragma-linguistic constructions in concrete realization.
4.1. Analysis of True and False Report
As pointed out, reportings of this study are produced with the total effort to persuade the listener of their truth-value. We will find out the clues which classify a direct reporting as a “lie”, by focusing on surrounding linguistic tools, because the part of reported speaker’s statement only contains “uttered words”, presupposed by the reporter”3. We aim to provide an overview of reportings which have been said originally and those who are only designed creatively as “truth” for personal benefits or other convenience purposes, but are not or just a little related with reality.
Let’s see three RS parts from corpus of total 31 conversations:
(1) S1: Emm s’po di a met tregu…mos e merr personale, garant s’u ni mir a najsen, po Erika ka fol keq për ty. Nashta s’e ka menu po tha që je valltarja ma e keqe e ma qesharake, edhe ka qef me kesh me ty. Po mos ta nin zemër menimi i saj s’o realiteti jot. Po ajo edhe ashtu e pime u kon
Emm… I don’t know if I can tell you…please don’t take it personal, she probably wasn’t feeling good, but Erika spoke harsh words about you. Maybe she didn’t mean it. She said you are the worst and funniest dancer in the world, and she loves to laugh with you. But, don’t worry sweetheart her perception isn’t your reality. Also she was drunk in that moment.
Declared truthfulness: TRUE Persuasion: 80%
When the reporting part contains hurtful words, as we see, it has some hedge constructions before transmitting the words of the original speaker. The construction: Erika spoke harsh words…is a generalized report or a preparation, which then gets deconstructed, when Erika’s exact words are transmitted: she said: you are the worst and funniest dancer in the world.
According to Knapp et al. (1974) , deceivers exhibit more uncertainty, vagueness, nervousness, reticence, dependence and unpleasantness than nondeceivers. But, in the case above, we can see that the speaker may be nervous due to recipient’s feeling, emotions that might be caused, or while talking, S1 may be uncertain if she took the right decision on showing these words or if she should continue/stop to tell the whole story: don’t know if I can tell you... This case expands social meaning of these uncertain words, which as we see, do not always determine truth-value. As DePaulo et al. (2003: p. 80) say, “truth tellers may express self-doubts, claim they do not remember things, or spontaneously correct something they already said, whereas liars would scrupulously avoid such admissions of imperfection”. The utterance after RS: don’t worry sweetheart… serves as a shield from behind to console her friend. S1 also reasons the original speaker by confirming that she wasn’t sober while talking, thus, wants to suggest that the offenses mustn’t be taken too seriously. Moreover, the teller doesn’t show a nuanced with self-interest information and her focus remains on empathizing with the hearer and paying attention on reported speaker’s emotional state, which may be relevant while interpreting accurately his/her intention.
(2) S2: Lena osht e ditne që ti s’din matematik, po menoj…për qato e niva dikon tu thon Lena s’ka tru. Sigurt, ta garantoj që sinqerisht folën edhe krejt u pajtun me to, u dok e sinqert, pom kujtohet, kah folke. Tani pom kjtohet që krejt kan kesh edhe njoni pej tyne tha: ajo osht krejt e pavlerë. Osht shum e ult pej teje me toleru ksi sene.
Po tlutna maje nveti, mos mle mu thirr shpiunk, po ta tregoj për tmirën tande
Lena, it is known that you perform really bad in math, I mean…that’s why I heard someone saying: Lena doesn’t have brains. Surely, I guarantee you they were sincere and I remember so many people agreeing with it and she looked very sincere, I remember, when she said that. I also remember that everyone laughed and one of them said: Lena is totally worthless. It is disgusting from your part to tolerate such labels. But please keep this to yourself, I’m telling it for your good.
Declared truthfulness: FALSE, Persuasion of hearer: 50%
S2 since the first comment, wants to orient Lena about her weak performance in maths as “worldly known fact”, thus to appear natural and rational issue, just in the beginning of “story”. General statement with no reference, by just defining a fact as “known” is very poor resource to take seriously, even though it may function successfully on several contexts—if the hearer isn’t curious about the author of that offense which actually is “converted in reality”, but focuses on reported utterance: as a de dicto interpretation (see Coulmas, 1986 ). As the author isn’t specified (the reported speaker is anonymous), the utterance can be defined as hearsay, which as a source, has the lowest possibility to be true (Willett, 1988) .
S2 dodges and misrepresents the facts quickly. The author of the same utterance, once is mentioned in singular (she said that), then the personal pronoun is in plural (they were sincere). Thus, S2 doesn’t intent to ruin any relationship of Lena with specific person, as she doesn’t pay too much attention to the producer of harsh words. Her intention clearly is to convince Lena about her inability to learn, by emphasizing the “absence of brain” and simultaneously to attack her self-confidence by “known facts” about “her mediocrity”.
Newman et al. (2003) suggested a relationship between content of communication and its style. According to their research (p. 666), “at least three language dimensions should be associated with deception: 1) fewer self-references, 2) more negative emotion words, and 3) fewer markers of cognitive complexity”. Two conditions are valid in our case, except the last—as false reports in our study contain more cognitive verbs as: I think, I mean, I remember, that serve to waste time while producing imaginative scenarios and also are tied with marking shades about their subjective assessment. Specifically, the mental verb I (don’t) remember helps speaker to confuse the hearer and making the utterance appear more sophisticated with mixed “facts”, which are reasoned by “lack of good memory”. Hereby, memorized moments, are employed without coherence: I also remember that everyone laughed…
Knapp et al. (1974) after analyzing seventy-six videotaped interviews to detect differences between deceivers and non-deceivers, hypothesized that “liars avoid statements of ownership to dissociate themselves from their words”—that’s why S2 often mentions many “authors”, although emphasizes the condition not to verify RS’s veracity. And by saying: It’s disgusting from your part to tolerate such labels encourages her to react angrily, start to assume/doubt who made such statements etc.
(3) S3: Elita, vllau i Tonit që jeton nGjermani folke për frizurën tonde. Edhe…Mos i trego atij ose Monikës që pot tregoj, po tha: Eli osht qesharake, si djal me ni stil palidhje. Kshtu që ti dokesh si djal edhe kan qef me kesh me ty. Pajtona me to që flokt i ki tpalame, dhe… po jo që ski inteligjencë. Po menoj…që dokesh si djal, se u habita
Ermmm… Elita, Toni’s brother, who lives in Germany commented your hairstyle. And…don’t tell him or Monica (his girlfriend) that I told you, but he said for you: she’s so funny and like a boy, your style sucks. So you look like a guy and they love to laugh with you. I agree with him that your hair is unclean and… but not about your non-intelligence, I mean…also that you look like a boy sorry…I got confused
Declared truthfulness: FALSE Persuasion: 0%
Declaring that the reported author mentioned is away and the contact with him is difficult, plus the warning “don’t tell him that I told you”, as the option to confirm it through reported speaker wants to be completely canceled—is enough material for the hearer to “sniff” that something isn’t fitting well with the speaker. It is like a warrant to urge the hearer not to try to find any accessible, verifiable source. She wants to make sure that persons that are close to the reported speaker won’t be aware of the reporting. The connectors, edhe (and)…serve to waste time while twisting the facts within context. As an excuse, she “self-regulates” her sayings, with the explanation that she has been confused (I mean…, I got confused) either earlier, or in that moment—depending what serves her intention more. The part she loves to laugh with you is free-direct reported speech, which involves the author on expressing original speaker’s attitude.
In 2nd excerpt, there wasn’t any specific “original author”, because purpose was hidden in convincing hearer about her “inabilities”, while in this case, S3 gives name of a mutually known person, Toni, whose brother is “the author”, what implicitly indicates tendency on creating negative perception about him. As we see, S3 cannot keep her story straight, as she includes many insults said by “Toni’s brother” and interpreted by her alone in incoherent manner, what may be perceived as personal opinions: you look like a guy. This shows that “lies vary markedly in the goals they serve” (DePaulo et al., 2003: p. 77) , as peculiar “main topic” chosen to be the focus, becomes associated with emphasized parts of RS, which also implicate liar’s purpose.
4.2. Typical Parts of (Un)True Reports
After scrutinizing the corpus, we identified three typical parts of RS, when speaker doesn’t only seek to quote a person, but s/he also displays intention to persuade the hearer and evaluate “utterance’s real intentions” that aims to convey. “The standard structure” they use to prepare the dialogue is constituted by:
1) Shielding/hedges elements, which prepare the hearer about the potential effects that the RS may cause, dis/encouraging elements, whose intention is to console the hearer (not) to take the offenses or appraisals seriously and also, to create a logical line that reflects objectivity or warn an unusual story.
2) Reported part is formed by reporting verb + the saying that is attributed to another person than the speaker, who aims that the hearer believes that s/he’s communicating original speech, if there isn’t any marker which expresses overtly the insecurity of authenticity of RS.
3) Speaker’s assessment about the speech—this part contains personal attitude of S about the overall speech reported—the suggested epilogue, which is likely to persuade and influence hearer’s actions, opinion towards reported speaker or him/herself (according to S’s intention). There may be included distorted interpretations, depending on the source of information, which “can lead different people to divergent assessments of the information” (Fitneva, 2001: p. 404) , if it’s not made clear and details are unverifiable (see Nahari et al., 2014 ). When the source it’s not entirely specified (e.g. saying a professor, a girl etc.), liar sees it as a paved way for more manipulation and adjustment according to his/her intent, as s/he knows it won’t be any address to be confirmed, nor revealed (see Figure 1).
These parts aren’t projected, pre-compelled, nor activated automatically when someone decides to convey words of others, as the hearer may understand the speech reported only through reporting part. Thereby, these additional phrases which surround it, show speaker’s concern to the upcoming attitude/response of hearer. They are usually ordered tactically on this chronological manner, because key reactions such as: anger, frustration or other extreme actions, that derive through these parts, may be triggered or prevented just before or after transmitting someone’s words. Thus, they are guided by psychological perspective: recursive awareness about reactions that may be caused and how the speaker thinks that it would be the epilogue according to his/her knowledge about hearer’s personality, strengths, weaknesses etc. Thereby, based on repeated (normal)politeness patterns, s/he recognizes that there are some conventional or knowledge-based expectations from the hearer, to not be very hasty on “hitting the point”, rather choosing to conform some pre-rules, operating through “hedges”, before going in the main part.
Shielding parts, based on their factually proved euphemistic/hedging functions, may be defined like a gift of language to people. They are preferred in normal relations, especially if S is going to transmit something very shocking that may generate tension, trigger the hearer to act furiously etc.
![]()
Figure 1. Parts of true/false reported speech according to their functions.
5. The Difference between Untrue and False Reported Speech
In two cases, respondents declared reports with “mid-truth” value. These include hyperbolized statements, with additional false effect or details. That is “to lie while saying the truth” (Falkenberg, 1982) or “telling the truth falsely” (Ekman, 1992: p. 37) . In this line, in communication, there are several statements that entail both components: creative part of the speaker (intentionally inserted) and the truth basis (factually proven).
Though comparable and similar, there is a conceptual distinction between false and untrue reported speech.
We view false reported speech as an outright lie. E.g.
a. Trina said she will attend the meeting
This is false reported speech when in reality, Trina has not even discussed the topic of being present in the meeting or she firmly has claimed that she won’t be there.
On the other side, the untrue reported speech is cognitively demanding, requires more effort and emotional investment in order to prove its entire veracity, thus evoking some suspicions about the degree of truthfulness. “If I said we went to the movies and you took it that ‘we’ was intended to include myself and my wife, then if you asked her and she said she did not, you might take it that the assertion ‘We went to the movies’ was false, though, perhaps, ‘I went to the movies’ was true.” (Sacks, 1975: p. 63) .
We can take this example:
b. The president says that he is against abortion.
This reported speech is simply untrue when factually, the President endorses some exceptions (rape, incest, health of the mother). In such cases, the utterance hasn’t a decisive property and “statements are received with skepticism” (Vrij, 2008: p. 48) . Following this, Marsili (2014: p. 157) , who calls these propositions fuzzy lies, states that “to say that the speaker believes that a proposition p is not true is to say that the speaker believes that the truth value of p is x, where x is 0 ≤ x < 1. By contrast, to say that the speaker believes that a proposition p is false is to say that the speaker believes that the truth value of p is 0”. These are graded truths, dodged assertions. To conclude, the untrue statement is more masked or vague, while a false one is a totally imaginative claim.
6. Discerning Veracity of RS
Summarizing the “shape” of RS turns in conversations, we can claim that it takes a special attention to individualize cues that show someone’s sincerity on reporting statements, as it is not found any conclusive/sharp evidence without errors, or fully accurate technique on diagnosing that—though many methods have proven success and received support on researches e.g. The Strategic Use of Evidence, imposing-cognitive-load, asking-unanticipated-questions (Granhag & Hartwig, 2015; Vrij, 2018) , The Evidence Strategy, which lists evidence then asks how they explain it, without mentioning accusation itself (Moston & Engelberg, 1993: pp. 228-229) . However, “many professionals are taught all sorts of techniques with no evidence that they actually work…and misjudging deception can have severe and costly consequences” ( Vrij et al., 2019: p. 303 , see Kassin et al., 2003 ). Nevertheless, if clues were overtly and universally recognizable, there wouldn’t be any false RS, because who wants to waste their time, just to show s/he is a liar?! The reason false statements exist, means that they can function as “truthful” and they do. Thereby, based on our corpus, we’ll give some of typical features of the pragma-linguistic and discursive cues of true and untrue RS.
6.1. Characteristics of True Reports
True reports, in general, do not connote a concise personal positive/negative position, meaning that the focus basically is on the accuracy of reported speech (remaining faithful on message transmitted), in the way which it emotionally may be perceived by the hearer or the manner which accidentally can be viewed: if it might be misinterpreted as attempt to disguise/hide intentions or other relevant facts. Thus, true tellers pay attention on the information and they don’t mind expressing doubts about maintaining full accuracy on transmitting exact words, but they generally are focused on core message. Their assessments are not very compelling and provide nuances only relevant with the statement reported i.e., they do not exploit RS as strong evidence to reach interests out of topic.
Below are summarized main features noticed after analyzing true reports in our corpus.
Express their insurance overtly. True reports are filled with uncertain epistemic modals in the assessment and the shielding part of reported utterance: I am not sure what she meant, probably she acts like this when she’s angry, it doesn’t mean it is real4, but the RS part is quoted with clearance without many “technical lapses” or different overall sense of it, although emphasizing chances of possible mistake while quoting (usually synonymous): she may have chosen another word; I am not sure…; that’s how I remember… I may have misunderstood, meaning that they safely state that potential emotional forces may be caused wrongly. Thus, truthful accounts “are not defensive about admitting their uncertainties” (DePaulo et al., 2003: p. 80) . That doesn’t mean that subjective assessments are absent. They just are more marginal and do not tilt only on totally “good” or “bad” implication with reinforcing devices, rather leave room for potential misunderstanding by their side and don’t show too much interest if the hearer is perceiving them the same way.
Spontaneous mistakes. It is clearer that sincere reporters don’t bother too much on gaining validation, nor hold a strong burden of what the hearer may think about their sincerity—thus, there is not too much effort to convince the hearer that RS has 100% credibility and that hearer must be fully sure about truthfulness of assessments included. It is fairly easy for them to emphasize potential misinterpretation spontaneously, what creates the openness for alternative options of evaluation, substitute words spontaneously, to not become hasty on judgments etc. On the other side, they convincingly hold their ground about the general informational characteristics of the main message heard: Though she surely said something like this…; that’s how she expressed herself, she was definitely tired, etc. Moreover, they accept responsibility about the claims they deliver, without including “third persons” as potential interpreters or by hiding reported speaker’s identity.
Shielding elements and brief assessments. True RS have many linguistic hedges in order to not compel any subjective mistaken implication. Assessments are more logical, than filled with subjective adjectives which are in lesser degree: it means that she wasn’t aware about it; he didn’t remember well; she was based on hearsays; don’t worry she wasn’t present on the party..., meaning that they are not “concerned with making a credible impression” (Vrij, 2018: p. 301) . They express doubt in shielding and assessment parts about the “real intent” and do not over-exaggerate intruding on the way the hearer views the negative side, in order to accomplish certain goals (Granhag & Hartwig, 2015) . Thus, shielding elements used aren’t determined on “persisting in a certain conclusion” or fabricating a scenario that would fit the speaker.
Details are related with location, time, emotional state. This is a valuable and consistent clue on our corpus. True-tellers give enough space to “situational” data and reported speaker’s state while were listening the words they report about: I was just tired from school and she came to tell me, she was drunk, right before the school break…, in the hall of the shopping center, right in school’s yard.
“Truth tellers’ strategy is to tell it all and to give as much detail as they can remember” (Vrij et al., 2019: p. 307) . The thematic nature of details is basically filled with circumstances on which the conversation occurred, emotional reactions that may have influenced the words chosen: she was nervous about the mark on math…, he just broke up with her girlfriend…, it was right after the fight, she is in her menses… Generally, “individuals who are good at sending basic emotions are consistently judged by observers as being more truthful” (Riggio & Friedman, 1983: p. 914) , and also the information about relevant details like spatial, temporal details are common for truth-tellers (Vrij, 2008) . “The ability to give details about events is commonly associated with having had a direct experience of those events” (Galatolo, 2007: p. 207) , while lack of them, together with inconsistency, concealing information and absence of evidence, are cues of deceit (Vrij et al., 2019; Granhag & Hartwig, 2015) .
6.2. Characteristics of False Reports
When the RS is false, the speaker is more concerned on his/her intentions that wants to achieve than on being coherent and reasonable in the eyes of the hearer. Thus, lie is more deliberate and “requires mental effort” (Vrij & Ganis, 2014) to be designed with reported part and composed with the implication of the purpose on making it, as usually no one wastes energy or jeopardizes his/her reputation just to test his/her abilities, without any benefit. Liar, as argued, aims to reason/adjust the statement in a way that speaker presumes it is justifiable by the hearer, and completely “natural” to be perceived as real.
Too much personal assessment. “Keep an ear out for generalizations and estimations” (Meyer, 2010: p. 79) . False report’s assessments, generally, in their core incorporate the “real” intention of transmitting RS, so that’s why the focus is on trying to convince the hearer to approve indisputably their attitude: is very impolite to choose her discourse; I don’t know why you hang out with her; she’s a bad company; Elira’s reaction was strange, you see? She has the courage to give you offensive etiquettes. E.g. calling someone strange doesn’t necessarily mean offense—so the liar is indicating incoherence between the adjective’s meaning and evaluation. Thereby, liars are fully committed on making clear their attitude and insist to make a point in their arguments trying to gain hearer’s approval about their “accurate assessment”. This clue is an excellent indicator and the main symptom noticed on our corpus. Claim by Newman et al. (2003: p. 666) that liars use “more negative emotion words”, is in concordance with vast majority of our results. But, also when positive evaluations are used, they are highly compelling and accompanied with deontic modals, tag questions: he said for you: she is my dream girl, isn’t she cute, everybody must marry such romantic guy, you must believe what she says; you have to follow his instructions.
In addition, they present themselves as the only reliable/objective reference and show commitment on persuading the hearer by pronouncing less the phrase “I forgot”, but simultaneously describing with evasiveness the “exact” reported speech e.g., she wants you to be guided only by her, so she said Linda isn’t disciplined enough; he always aims to dominate, so he claimed: no one is looking for me. This tells that liars “aim to mask their attitudes” (Meibauer, 2018: p. 362) through twisting declarations. Showing too much their assessment about other’s statement and persisting on its truthfulness, overtly shows subtle differences with a true report as more neutral and impartial connotationally. The focus on the subjective assessment of the report and insistence to believe liars’ evaluations about reported speaker’s intention, is one of the consistent signs of a false RS.
Giving “some” details. “Liars’ psychology is driven by planning, strategizing, and calculation” (Vrij et al., 2019: p. 300) , so there aren’t many opportunities to be disapproved. They deliberately seek to control and appear consistent in every possible component and chance that the hearer has, not to think otherwise than him/her, by overpreparing details (though with thematic incoherence as was emphasized). Dissimilarly, in the study of DePaulo et al. (2003) , the most reliable indicator in size effect d is that liars are less forthcoming for details than truth tellers. In this line, Vrij et al. (2019: p. 299) claim that “offering details that sound plausible requires imagination, and liars may lack such imagination”. However, places they choose to fill with details, in our corpus, are different from true RS, and don’t provide sufficient evidence about actual features of the utterance transmitted, rather have to do with personal (unnecessary) data: (she lives in Germany, she has a dog), introvertness/shame of reported speaker etc. In other words, they are irrelevant with the statement reported, but relevant with their intention, and persist to not be revealed through potential aim of the hearer for confirmation, from the “fear that investigators will check such details” and discover the lie (Vrij, 2018; Nahari et al., 2014) : She lives in Germany; she’s away from here; he wouldn’t admit to respond you; he acted like he’s unconscious, she hates to ask her about the past.
Liars make efforts to remember earlier statements (Vrij, 2008) , but in our case, they fail to achieve it, as during the talk they mix facts, avoid questions and protest or protect themselves if they are contested. “For a lie to be believable, it has to contain sufficient details to bear the characteristics of a self-experienced event” (Vrij et al., 2019: p. 299) , so they try to fill the gaps of certain implications of lying. But as we saw, their creativity betrays them, thus, they cannot keep coherent chronology of events, neither thematic nature of details.
Reported speaker’s “awkward” personality. Though their words’ meaning is “confident” (without including face and intonation), main part where the liar expresses doubts is the identity of the reported person and his/her emotional state when speech was (falsely) uttered. The original author of offense words is rarely, if not, mentioned, just not to associate the hearer with the “original author”, but only with “his/her” statement. Thus, the source of RS is of low credibility: very vaguely described with vast of interpretations, not always fitting the assessment given, because they generalize reported speaker’s personality and attitudes.
So, in order to fit with their intentions, first, liar describes confidently embedded attributes, then goes: she always talks like this about you…in order not to take as something unusual, and tend to use more negations and more generalizing terms such as always, never, nobody, or everybody (Vrij, 2008) . They avoid suggestions like: go and ask her, rather they manage it by some moral rules: by implicating that if the hearer wants to prove veracity, s/he will be labeled as a spy, untrustworthy, and want to provide the guarantee that nothing will be revealed with phrases: I gotta tell you something, but first promise me you’ll keep it to yourself. Thus, a relevant condition is also (non)description of reported speaker: by whom are the words transmitted, what does the hearer know about attitudes of the speaker towards the person mentioned as original author.
Even when they are clearly identified, the original speakers are far away or on higher social position and the liar knows it won’t be confirmed: Your boss said: Ema can’t finish any work without complaining, I’m planning to fire her—the intention here may be to frighten, “discipline” or make her feel unworthy.
“Feign” surrounding elements. Liars’ counseling words are kind of “weak” when their purpose is to attack hearer’s self-esteem or “good opinion” they have about the reported speaker (Ex. 2): e.g. while they specify the negative implications of RS: it is very hurting, doesn’t mean he’s an angel, they clearly make sure they urge the hearer to behave how they want: don’t ever talk with him, but stay away from her, So they use these strategies that “they think will look honest” (Hocking & Leathers, 1980) and wish to be seen as a victim of the “same mistake”. They make efforts to look friendly, positive even though they’re triggering nervousness, anger, “thus they are concealing negative emotions and falsely claiming positive feelings” (Ekman et al., 1988: p. 414) . They explain their “similar case” and their mistake for believing someone who lies them, to try to urge the hearer not to act in such way. By sharing false empathy and evaluating the RS as “normal statement” for that “bad person”: It hurt me too once, you mustn’t be surprised by him/her. Their pity sounds as “cynical assertion” (Kenyon, 2003) , because many liars “do not have the moral high ground, the emotional investment, or the evidentiary basis for staking their claims” (DePaulo et al., 2003: p. 77) . They tend to avoid additional questions, because their creativity may betray them and get caught, or it is possible to foresee potential questions, in order to prepare their answer and erase the doubts by “imagined occurrences: (Vrij, 2018) : if you are interested about other classmates, they weren’t there. The fact they didn’t have the reported person right there, gave them more freedom to explore imagination, but (non)verbal fillers were present: mmm…. ermmm…, when the response was to an unexpected question.
Different versions within a report and generalization. Liars mix many details (often skillfully,) in order to make it complicated for the hearer, and indirectly giving permission to themselves to have the final say/explanation. This is commonly trustworthy sign e.g. when the facts are inconsistent or neglected is “also evidence for the prosecutors to doubt reliability of informant’s testimony” (Garrett, 2011: p. 123) . So sometimes is not that they cannot manage their story appropriately/coherently, but want to make it look complex and they continuously change statements “in order to make it fit the evidence presented” (Granhag & Hartwig, 2015: p. 237) .
Summarizing, if the lie contains many details, some clues are argued to be unpreventable and difficult to be concealed, no matter how cleverly they are prepared. But what happens if liar faces unexpected questions that reveal inconsistency in events? The respondents who didn’t ask for much time to prepare the false report, mixed many details, which would confuse the hearer in that measure that it would be even hard to formulate reasonable questions: we were at the park…and while she was lying down in her bed; Many parts had ambivalence e.g. when for the same person is said: she doesn’t love anyone, then: she is in love with him. Thus “liars appear to be less immediate, more ambivalent, more uncertain, and their statements are less detailed and sound less plausible” (Vrij et al., 2019: p. 304) and these flaws must be cleverly taken into consideration.
Instead of thinking about real facts of the past, they orient their memory on creative resources to fabricate it to look “real”. However, strategies aren’t totally creatively constructed, but their base is rooted on previous reactions of certain persons or the roles they fulfilled e.g. if a strategy has succeeded in the past, then will be usable, if not, the liar would evade it.
They may appear sincere first, but if we analyze deeply the kinds of persuading efforts, we notice several hints that clearly show that S wants to do something for her benefit, not just to inform the hearer.
7. Pragma-Linguistic Differences and Similarities
We distinguished several discursive and linguistic features/tactics that offer guidance on detecting if a reported speech is valid on the aspect of truth-value. “Linguistic components, such as prosody, syntax, and the lexicon, must be taken into account when investigating the linguistic structure of lying” (Meibauer, 2018: p. 358) .
By the measure and significant difference, we’ll see which of them can offer indirect instructions on anticipating the discourse modelling of a false/true reported speech5 (Table 1).
In Multivariate MANOVA, when variables were considered jointly, there was also significant difference: F (11, 19) = 122.64, p < 0.0001, eta2 = 0.98.
This significant and clear difference on these pragma-linguistic elements in this high level, may have resulted due to efforts of respondents to manifest accurately, in strict sense, “how a lie would be modelled like”, as the reports weren’t totally spontaneous, but respondents were urged to use markers that index their insincerity on their everyday life when they produce lies.
Indefinite and general nouns which do not specify the person are common in different parts of false RS: some students, a professor, a very high position person (to make sure he won’t be contacted)—which is initial indicator of vagueness/weakness on facts presented. They function as “the main source” of information in 10 conversations, and are used 73 times within 17 of them: a woman who works with her told me…, a student heard Elita shouting your name and…a stranger in café laughed with your hair…, by trying to shift the focus on the saying and its effect, which the speaker wants to accomplish by convincing hearer about an opinion she wants to say herself, but chooses to do that through someone, so doesn’t hold the responsibility of “inappropriate labels”.
“Liars don’t want responsibility, so they replace the pronoun ‘I’ with ‘you’” (Meyer, 2010: p. 82) . In our study, in 13 of 17 false RS conversations there wasn’t any specific author and the source of information in many of them, despite a stranger person as initiator, is “they said”. In addition, the few references that exist in deceptive reports, are people who cannot be contacted easily by the hearer, or are distant socially, as “liars report more unverifiable details than truth-tellers” (Vrij, 2018: p. 310) . Pronouns he/she are more used on the assessment part, where liar is committed to compel opinions for someone. Moreover, false RS are equipped densely with personal assessments (with more than one sentence) and most of them (70.5%) intend to compel negative evaluations about the reported speaker.
![]()
Table 1. Statistic difference of pragma-linguistic cues.
In deceptive RS, connectors dhe, edhe, e (synonymous with and) were repeated mechanically by trying to lose time on using/relating creative “facts”. Thus, “successfully lying to another person usually involves the manipulation of language and the careful construction of a story that will appear truthful (Newman et al., 2003: p. 671) .
True reports contained significantly higher number of advices not to take personally and a specific, contactable author, who was mentioned 58 times in overall corpus of true RS, without momentary doubt or being “confused” by jumping once in singular then in plural. That’s why names are in definitive form (in Albanian are expressed through suffixes: vajz-a, djal-i, (the boy, the girl). They mostly didn’t mix details, nor “overflowed” the discourse with personal evaluations.
Epistemic/Evidential Modals
In this subsection will be considered statistically the presence of markers that tell likelihood of accuracy on transmitting exact words, intentions of the reported speaker, or evaluations by the reporter, regarding the veracity of RS. Evidential and epistemic markings as “rhetorical strategies” (González et al., 2017: p. 69) were highly present in respondents’ discourse, especially at the “shielding part”, when the speaker tries to avoid the literal interpretation of the offense or, contrastly, tends to make it seem ruder or more certain than actually is. “These markers are pivotal for the analysis of lying” (Marsili, 2014: p. 164) , more specifically, is relevant in which part of the RS they are inserted.
Modal words, which construe the meaning in terms of possibility and predictability of the action described, are used by both parts, but not in the same distribution (Table 2).
Utterances with low certainty epistemic modals before RS were higher in true reports, while deceptive ones are highlighted with more “confidence” linguistically in the reported part. But this confidence doesn’t fit with other paralinguistic cues e.g. including smile while talking, or unsustainable intonation.
In paralinguistic level, Ekman & Friesen (1982) , Ekman, Friesen and O’Sullivan (1988) made distinction between felt smile, “when the person actually experiences positive emotions”, and false smile—“when deliberately made to convince another person that positive emotion is felt when it isn’t” (Figure 2).
![]()
Figure 2. Respondent’s smile while thinking “creatively”.
![]()
Table 2. Distribution of epistemic modals.
In addition, intonation’s affect is an important clue in understanding speakers’ sincerity. When an untrue report is delivered, the vowel is longer, perhaps as a tendency to activate imagination which requires time, or to show more emotional impressions. “The intonation implicates false positive emotions: pit, sympathy, compassion, and the tone and some gestures may leak true feelings” (Ekman, 1992: p. 40) . But on the other side, stress, nervousness is experienced tensely, thus, harder to be controlled consciously, because of the fear of being caught. Although their confidence is mostly simulated, their verbal cues denote certainty, which is not manifested on their speech coherence, neither on their tone of voice
Liars struggled on meeting their intentions which were mainly and carefully hidden on assessment part, thus their weapon was to show assurance typically on that part, in order to impose and “load” hearer’s mind with “true purpose of past statement”, while they “didn’t have enough evidence” to identify the author, who in cases of untruth RS was either of high status (to push away the possibility of verifying), away from Prishtina, or a totally stranger identified only by his profession, so they don’t generate unwanted outcome.
8. Discussion
In this study were examined connections between lying and the structure of a language, precisely pragma-linguistic and other psycho-social characteristics of (un)true reported speech, when the speaker transmits a statement of a commonly known person with co-interlocutor that expresses certain attitudes, humiliation, or compliments about the hearer, whether falsely or not. Even when RS was false, they argued that are able to encode totally imagined constructions that fit their agenda, and are not easily recognizable, because in 8 of 17 conversations interlocutors were convinced more than 80% that speaker was sincere. However, many linguistic/discursive devices were shown with significant difference on their distribution at true/false reports, indicating that usual general tendency of liar is indirectly noticed through language.
“Lies are necessarily more difficult to construct than truths” (DePaulo et al., 2003: p. 79) , because when imagination or minimal evidence are the main source of providing coherence and logic of a false story, it may betray them during descriptions of details with different versions of circumstances in different parts of the story.
Motivation of decision to lie may be related with self-or group interest (favor-gain, or speaker may feel insecure to do that), just for fun, feeling more powerful by making someone believe something (ego’s satisfaction feel smart enough to persuade, as additional benefit), to challenge/protect themselves from any embarrassment/punishment. Their initial strategy to overgeneralize and giving primacy to personal assessment are primary determinants of untrue RS and extremely suggestible cue of untrue RS. They also insist to make sure the silence of the hearer by indirectly threatening with accusations as calling her spy, cutting off friendship etc.
9. Conclusion
It is shown that truth tellers are more neutral and comfortable on attitudes they share towards the reported speech. Their shielding utterances are accompanied with epistemic modals: maybe, probably, possibly. In the other side, liar’s attitude is more categorical in defining the intention of the reported speaker. Even in the absence of the clear statement in the reported speech, they don’t hesitate to play the role of a “persecutor” of sayings, like their impressions are indisputable truth. They give more space to the assessment, even if it results in total discrepancy with the speech.
Regressing at modals’ usage, as it is argued, true reporters involved more words that indicate uncertainty in the shielding and evaluative part, while they expressed more certainty on the information they were conveying through reported speech. However, 37% weren’t totally sure if they were selecting entirely exact words of the past, but all of them held their ground about the core of message. Though cues sometimes may be sporadic and subjective impressions, the ones which maintain some regularity on their manifestation gain relative validity. Thereby, we suggest that the hearer must be vigilant to gather information and to show clearly suspicions without feeling guilty on embarrassing the hearer, because it is in his/her right to prevent being called foolish.
Acknowledgements
I thank all participants of this research (many of them my colleagues and cousins) not only for being part of study, but also giving me valuable suggestions on eliciting prominent cues.
NOTES
1Nevertheless, in this article we won’t deal with the scale of fitting by comparing it with the original formulation, because we do not even have access to reported speaker or the original situation to define its originality, but we will pay special attention to the (non)verbal cues and other implicit assumptions after the speaker tells if s/he was sincere or not.
2Nonverbal cues didn’t show any strong supporting evidence on telling if someone is lying. E.g. the study of DePaulo et al. (2003) according to (Vrij et al., 2019: p. 302) “has discouraged many researchers from examining nonverbal cues” (see also Riggio & Friedman, 1983 ; Kraut & Poe, 1980 ). Many of clues like displaying gaze, smiling etc. were overestimated.
3The addressee may judge the truthfulness only based on the reported part: e.g. by tone of voice, as subjective impression of (in)sincerity, the (im)possibility that e certain speaker may declare something, based on episodic memory etc. But, for an analyst is difficult to be based on the saying and subjective evaluations, as she/he is not aware for many contextual needed information/details.
4Examples are only written in English.
5We must emphasize that false RS’s utterances were longer than true ones, thus, contained many repeated and generalized words (all of them were counted).