Crossing Linguistic Bridges: Unravelling the Art of Persuasion Among Albanian Writers in Greek


Intercultural communication can be seen as an act that involves an exchange between two or more persons from different cultures, either via a common third language or by means of a linguistic code which is the L1 of one of the persons involved. As an area of study, intercultural communication also investigates the quality of interaction and, in particular, the influence of culture on the success or failure of the act. Since communication is usually materialized through the medium of language, intercultural communication, as a study field, is also discussed under the umbrella of applied linguistics (discourse analysis).

While PML failure is clearly also intracultural, most interest has been placed on its intercultural variety as this situation is more demanding when speakers of an L2 are required to understand (decode) but principally produce (encode) meanings that are not only structurally correct but also culturally acceptable and suitable to the situation. Research in which communication breakdown takes place through a third medium (a third common language between communicators), which occurs more frequently with international languages, has been examined under the term Lingua Franca; i.e., pragmatic failure in non-native/non-native encounters (see, a relevant study by James, Scholfield & Ypsilandis, 1994).

While the mastery of the structural parts of an L2 can be achieved either after certain dedication and effort or perhaps acquired in a more natural fashion, conquering the L2 pragmatic level presents considerable difficulties. This may be due to the fact that while structural rules are descriptive and can be practised, learned, and used straightforwardly, pragmatic conventions are fuzzier, and communicators’ intentions are not always suitably encoded for the situation (phrased in a manner transparent to the recipient), or when these are decoded in a different way than the one intended or expected.

Several studies have shown disagreements between learners' pragmatic and grammatical development, resulting in cases where even the best L2 learners commit pragmatic infelicities when communicating with native speakers (Bardovi-Harlig & Hartford, 1991, 1993; Derakhshan & Eslami, 2020; Omar, 1991; Takahashi & Beebe, 1987).

Scholarship application letters are an area of an essentially authentic communication occasion requiring specific language competence and are, therefore, an appropriate choice to investigate the pragmatic perlocutionary force of persuasive strategies through the variable of suitability, as these strategies involve inducement for the participants' self-promotion (from here on also applicants/authors). Unsuitable employment of such strategies has been abundantly documented in a series of articles (James et al., 1992, 1994; Ypsilandis, 1994; Psaltou-Joycey & Ypsilandis, 2000; Papadopoulos & Ypsilandis, 2017; Farnia, Ypsilandis & Ghasempour, 2019), causing pragmatic infelicities in this specific L2 persuasive context.

It is, therefore, the purpose of this paper to investigate the pragmatic perlocutionary force of certain persuasive strategies in a setup involving intercultural communication of a minority language. Persuasive strategies in application letters for a scholarship, employed by L2 learners, were investigated in terms of their suitability to the task by recording teachers’ reactions via an internet evaluation tool. Independent variables, such as language level, age, and gender of the applicants, and educational level, teaching experience, and knowledge of L2s were considered for the teachers-evaluators, in addition to age and gender. The speech act investigated is of true value to both the participants involved (learners and teachers), as the former often find themselves in a similar situation, while the teachers typically react to analogous infelicities as members of scholarship assessment bodies. For this reason, the research design is realistic and valuable, as written persuasive discourse is part of every document a person would ever write to be read by another person (Frederick, 2012). Additionally, the selected speech act is regarded as a gate opener, often helping the applicants to advance their careers and to get educational rewards. Following this opening, the first section is dedicated to the discussion of PML failure. The research method then unfolds with a description of the design and procedure, the participants, and the tools used for the collection of data. The analysis of the results follows, and major findings are recorded. The discussion and conclusion are dedicated to an examination of the major findings and the limitations of the present study, with suggestions for future research.

Pragmalinguistic (PML) failure

The acquisition of pragmatic competence (i.e., the formulation of suitable meanings related to a specific situation in an L2) is one of the most difficult aspects of the L2 acquisition process (Bardovi, Harlig & Vellega, 2012), although there is a valid concern about whether pragmatic competence is teachable, given the complexity of pragmatics (Taguchi, 2013). On the other hand, this language aspect, also considered to be a skill, is of utmost importance to be mastered and has attracted the attention of researchers (Foster & Ohta, 2005; Kasper, 2000; Malmir & Derakhshan, 2020; Rose, 2009). This involves a) an applied linguistics research interest, focusing on the language use of the non-native speaker, and b) a more practical (pedagogical) aspect, which investigates the acquisition of L2 pragmatic knowledge (Kasper & Roever, 2005; Taguchi, 2019, 2020). Scholars look into how non-native speakers of a language acquire pragmatic knowledge, understand it, and use it in the expressed written discourse of the target L2 (James et al., 1992, 1994). Findings in these studies aim to assist the teaching of this language aspect and increase awareness and tolerance of the 'other'. In this light, Leech (1977, as cited in Thomas, 1983) states that through the employment of a suitably encoded utterance, carrying a certain pragmatic force, defining whether the speaker is polite or clear, the speaker gains many advantages when interacting with other speakers. Figure (1) below is created to schematically present the productive condition in which most research has been focused. The opposite dimension/direction (native speakers encoding messages which are to be decoded by non-natives) has not been equally researched.

Figure 1.Possible conditions in intercultural communication

It may be concluded that when speakers encode messages in a way that these are decoded by others as intended by them, and in the way expected by the recipients (mostly natives of the communication medium-language), there is no problem in communication (condition A). However, in the case in which producers/encoders of messages do not code these in a suitable and a situationally predictable manner (condition B), or from the opposite end when “the receiver decides and chooses to understand something different from what the speaker intended and if he or she failed to perceive the intended expression, then there is pragmatic failure” (Leech, 1977, as cited in Thomas, 1983; Riley, 1989; Thomas, 1983). This latter type of PML occurrence is added to the productive infelicitous encoding dimension and was recognized by Miller in 1974 (as cited in Thomas, 1983) as a receptive decoding side of conceivable misunderstandings, which the scholar claimed to arise during communication for the reason that receivers of messages often fail to understand the intentions of a speaker. Both possibilities lead to the pragmatic infelicities described as PML failure.

In this respect, failure may also be receptive, not only in terms of infelicitous decoding by the native speaker but also by the non-native learner’s inability to decode native talk. It may then be possible to claim that the problem initially lies a) with the encoder, and at a second stage, b) with the decoder, as the suitable and unambiguous coding of messages in a way appropriate to the situation is a significant skill to possess in intercultural communication setups. Note, however, that the native speaker possesses the advantage of selection by being linguistically equipped to converge to the non-native’s level/genre while being fully aware of what s/he is doing. On the contrary, for the learner, it is not a matter of a conscious decision as s/he is often not aware of the linguistic and social consequences of his/her sayings.

One solution that is often recommended would be for native speakers to increase their tolerance in intercultural encounters where their L1 is used as the means of communication. In addition, Leech (1977) provided a more linguistic explanation of pragmatic infelicity based on the polysemy of languages and argued that certain utterances could be interpreted as rewards, criticisms, and rejections, but also as a combination of the three (as cited in Thomas, 1983). In this direction, an example was provided by Thomas (1983); the question “Would you like to come and sit down?” which, depending on the situation, could be understood as an instruction, a request, an invitation, or something in between the three. In this example, the pragmatic force and thereon the implicature within these three possible interpretations differ, even in settings of communication among native speakers, not to mention between speakers of different L1s communicating through a medium that is the native language of one of the two.

The phenomenon of pragmatic infelicities in this specific persuasive speech act (writing a letter of application for a scholarship) has been explored and supported by evidence in a series of articles (Farnia et al., 2019; James et al., 1992, 1994; Papadopoulos & Ypsilandis, 2017; Psaltou-Joycey & Ypsilandis, 2000; Ypsilandis, 1994). The results of these studies recorded different types of PML failure committed in English, following the same research design and procedure, with several subjects speaking different L1s. In more detail, James et al. (1992) investigated the persuasive strategies by Greek learners of English and recorded the reactions of English natives with the PML failure suggested. In James’ et al. (1994), a list of the strategies from the 1992 study were judged by different non-native speakers of English (Poles, Portuguese, French, Germans, and Syrians), and differences from the English natives’ reactions were noted down (an instance of English as a lingua franca). The Psaltou-Joycey and Ypsilandis’ (2000) study followed the same procedure to investigate the reactions of Greek natives to the persuasive strategies of Russians and Spaniards when writing in Greek. A strong influence of the Islamic religion to the selection of persuasive tactics by the subjects became evident in a study by Farnia, Ypsilandis and Ghasempour (2019), which followed the same procedure for the collection of data, with subjects writing in Farsi and English. Applicants started their application with a reference to the Prophet and a considerable employment of ethos appeals was identified.

Other researchers who examined different speech acts have come to similar conclusions stressing the transfer element from the L1 conventions to the L2. A few examples are the following: a) Al-Mahrooqi & Al-Aghbari (2016) examined the British and the Omani ways of refusing requests. The study also concluded that transforming the rules and the cultural norms of the first language to the target language results in embarrassment, misunderstanding, or communication failure. Alrefaee (2020) also investigated refusals to offers and suggestions of Yemeni EFL learners and compared the findings with refusals of Arabic native speakers (ANS) and English native speakers (ENS). Pragmalinguistic failure again was claimed to result from L1 transfer. Saleem, Anjum, and Tahir (2021) investigated Pakistani English learners’ responses to apology speech acts and further supported the transfer hypothesis. It is evident that failure in all cases was attributed to L1 language transfer and violations of Grice’s (1975) maxims of politeness. This work wishes to contribute to this area of study with the provision of more data from a minority language (Greek).


Design and Procedure

The design of this work followed the exact same procedure scenario suggested in James et al. (1992), which was developed in four stages. It involved the writing of a letter for a fictitious scholarship in English by foreign nationals who were invited to submit it and present themselves in the most convincing manner, together with the reasons they thought they were eligible for the award.

Figure 2.Study Procedure

In this work, the design followed the exact procedure scenario suggested by James et al. (1992). The process unfolded in four stages. Initially, Albanian participants were requested to write an invitation letter to apply for the scholarship without receiving further instructions. Next, an initial evaluation of the types of strategies used by the applicants was conducted by two scholars specializing in persuasion. The strategies extracted from the letters were categorized according to the Aristotelian Ethos-Pathos-Logos framework. Subsequently, a protocol list containing three selected representative examples of each type was created for further investigation. In the third stage, Greek native teachers acted as judges and evaluated the body of strategies for their suitability in the context of the specific culture. The judges were provided information concerning the writing activity from which the examples were extracted. The evaluation was facilitated through a purposefully created scale-formatted internet tool, allowing teachers to record their reactions in a numerical format, thus offering more precise and accurate documentation of their opinions.

The research questions formulated and explored based on this framework were as follows:

1. To what extent were the specific persuasive strategies found suitable by the Greek judges for the particular linguistic situation (applying for a scholarship in Greece)?

2. Were the employed strategies different from those used by Greek natives when writing for an identical task in another L2 (English)? This question aimed to investigate the consistency between the productive employment of strategies by Greek natives when writing in another L2 (English) and the receptive perlocutionary force of these strategies and their suitability when used by Albanian learners of Greek. It focused on the selection of strategies written in an L2 and their suitability for the exact same task.

3. Finally, it was explored whether the new and tested evaluation instrument (see Tools, below) was appropriate to increase the sensitivity of judgment and offer a better understanding of the phenomenon.


The study involved seven native Albanians, learners of Greek as L2, who wrote the letters of application. Among them were six adults and one younger individual, representing beginner, intermediate, and advanced levels of Greek language proficiency. All participants resided in Greece. Additionally, 176 Greek primary and secondary school teachers, who were Greek natives from the public or private sector, acted as judges and evaluated the selected strategies.


A digital questionnaire, created via the Smartsurvey application, was utilized for the collection of evaluation data. Previous research on the topic typically employed qualitative analysis or Likert scale quantitative examination to evaluate and record the perceptions of native speakers (teachers in this study). In this work, a somewhat novel instrument was introduced and exploited for the task. This instrument was also tested for precision of judgment and sensitivity to illuminate the investigated topic and contribute to its understanding. The measurements were conducted using a fuzzy bar/rod (ravdos) suggested by Vougiouklis & Kambaki-Vougioukli (2011), which was then transformed into a digital continuous scale scroll bar by Katsarou and Kambakis-Vougiouklis (2022). Native Greek evaluators were instructed to move the scroll bar to the right based on their judgment, with further movement indicating a more positive assessment. Given the fuzzy nature of the construct of suitability, which is challenging to measure in a continuous ratio scale format, the Ravdos method was selected to record judgments in a fuzzy mode (scroll bar movement to the right), ultimately transforming the teachers’ opinions into a ratio format. While the teachers adjusted the scroll bar to the right in accordance with their judgment, the application recorded their opinions in the background using a ratio scale Ravdos ranging from 0 to 100. The collected data were then compiled in an Excel file generated by the application and subsequently transferred to the statistical software SPSS V.28 for data management and further analysis. The questionnaire comprised two parts: the first part included questions aimed at eliciting information about several independent variables, such as gender, age, education level, profession, professional experience, and L2 language background, to explore potential statistical associations with the teachers’ responses to the questions (dependent variables). Final data coding and analysis were conducted using SPSS V.28, a statistical program designed for quantitative data analysis.


In this section, the data analysis begins by presenting descriptive statistics of the independent and dependent variables, including counts, means, medians, and modes. Subsequently, relationships between these variable types are examined, supplemented by qualitative comments where necessary. The initial focus is on personal details questions related to demographic data collection. In the second part, suitability scores of the strategies are provided alongside statistical values, along with qualitative comments. This mixed-method approach combines quantitative and qualitative analyses to assess how specific persuasive discourse conventions were judged for their suitability in the given context. Additionally, associations among variables are explored and presented, particularly where statistically significant relationships are identified.

Descriptive Statistics: Independent variables

Gender. One of the survey participants was male, two were female, and four did not specify their gender. Out of the 176 teachers, 94.9% (n=167) were female and only 5.1% (n=9) were male. This perfectly represents the field of education setting in Greece (and elsewhere) in which the majority of professionals in this sector are female (Startz, 2019).

Age. Most applicants/authors did not declare their age except in one case, a 21-year-old. The age of the judges/reviewers ranged from 23 to 60 years (M=33.70, SD=7.88). More specifically, 38.6% (n=68) were between 23 and 30 years old, 45.5% (n=80) were 31 to 40 years old, 11.4% (n=20) were 41 to 50 years old and 4.5% (n=8) were 51 to 60 years old. As the questionnaire was posted online, it was reasonable that younger teachers would mostly answer the call, as they are more familiar with the new technologies and follow the community of practice networks where the questionnaire was advertised. When the age variable was tested for possible associations with the dependent variables, two negative statistically significant correlations were found with a) the logos first (r=-1,57, p= 0.38), and b) the pathos second examples (r=-149, p= 0.49), at the 0.05 significance level. This association revealed that the younger the age of the judges, the lower an evaluation score was provided.

Education level. 30.7% (n=54) of the judges were found to have a bachelor's degree, and 3.4% (n=6) of those had a second degree. 62.5% (n=110) had a master’s degree and 3.4% (n=6) had completed a Ph.D. It turned out that most judges had a master's degree, given that in recent years the skills required to acquire a teaching position in the Greek market have increased significantly, and this fact has positive benefits for this study as subjects appeared to be better equipped to offer more acute judgments.

Profession. 42% (n = 74) worked as primary school teachers, 11.4% (n=20) worked in secondary education, while 13.1% (n=23) worked in the public sector providing tutoring to children with difficulties. 33.5% (n=59) offered private tutoring.

Professional experience. The years of teaching experience of the judges ranged from those who had only a few months (recently entered the profession) to those with 37 years of experience (M=8.90, SD=7.48). Most of the 63.1% (n=111) were in the 0-10 year band, 31.8% (n=56) declared 11-20 years of experience, and 5.1% (n=9) were 21-37 years. The standard deviation was low (SD=7.48), which means that the dispersion of the mean values was not significant, and most subjects were in the middle of the Gauss curve. Therefore, considering the mean professional experience of the subjects (M=8.90) this can be considered sufficient time to accept teachers' judgments on the suitability of persuasive strategies.

Languages. Knowledge of languages other than the Greek language was also examined. 42.6% of the judges (n=75) claimed to speak English, 8.5% (n=15) English and French, 12.5% (n=22) English and German, 9.1% (n=16), English and Italian, 2.8% (n=5) English and Spanish, 0.6% (n=1) English and Russian, 1, 1% (n=2) English and Bulgarian, 1.1% (n=2) German, 0.6% (n=1) French, while 6.8% (n=12) claimed to be multilingual while 14.2% (n=25) spoke only Greek. It should be noted that an increasing number of teachers in service speak a second language in addition to English, with the percentage of multilingual being relatively high in this sample. Therefore, the majority of judges had the experience of learning an L2 and therefore possessed increased awareness of the difficulties one encounters when expressing themselves in an L2.

Descriptive Statistics: Dependent variables

As mentioned above, the pragmatic implementation force of the persuasive strategies was measured through the construct of suitability and examined under the three categories of Logos, Ethos, and Pathos. As the standard deviation was high in all the evaluation scores below (see, Table 1, Table 2, Table 3), the median and the mode scores, in addition to the mean values, are presented to offer greater precision in the scoring of the judges. The high standard deviation shows that the variation of votes was spread and not concentrated around the mean, suggesting that there was a notable difference in the scoring between the judges. This can be observed in the more detailed analysis described below.

Logos: Regarding the logos strategies, teachers judged the first and third examples (see below) to be of medium suitability (M = 51.65, SD = 24.93), and (M= 47.58, SD= 24.98) respectively, while the second example was significantly lower suitability (M= 21.19, SD= 19.74). Although the use of logos strategies was thought to be an advantage in scholarship application letters (Psaltou & Yspilandis, 2000), in the second example the applicant expresses himself in a direct (imperative) and demanding form (arguably due to a lack of knowledge of a more appropriate syntactic structure), a style of writing considered unsuitable by the judges (see, Table 1).

n % Min Max Mean Median Mode SD
Logos 1 12 26,6 0 100 51,6 52,2 47 24,9
Logos 2 12 26,6 0 99 21,1 16,00 0 19,7
Logos 3 12 26,6 0 99 47,5 50,00 29 24,9
Table 1.Descriptive statistics of logos evaluation scoresSource: Calculated by the author

L.1. "I would like to apply for the scholarship to learn Greek because I have lived in Greece for 20 years and it will help me in my work to better understand what is said to me." ("Θα ήθελα να κάνω αίτηση για να πάρω την υποτροφία για την εκμάθηση της ελληνικής γλώσσας, γιατί μένω στην Ελλάδα 20 χρόνια και θα με βοηθήσει στη δουλειά μου να καταλαβαίνω καλύτερα αυτά που μου λένε.").

L.2. "Give me a scholarship to learn the Greek language better" ("Να μου δώσετε μία υποτροφία, για να μάθω την ελληνική γλώσσα καλύτερα.").

L.3. "I would like to apply for the scholarship to learn the Greek language because it will help me better understand what is said to me." ("Θα ήθελα να κάνω αίτηση για να πάρω την υποτροφία για την εκμάθηση της ελληνικής γλώσσας, γιατί θα με βοηθήσει να καταλαβαίνω καλύτερα αυτά που μου λένε.").

The use of logos strategies does not appear to lead, at least in most cases, to PML failure (a collective mean of the scores for the three examples was M=40), nor was it considered particularly successful.

Pathos: As for pathos strategies, the teachers judged all three examples to be of low suitability, far below the middle of the scale (see, the mean, median and mode values in Table 2 below). Notice that the most selected score (mode) for all examples is 0, which shows that most judges found all pathos examples to be unsuitable while some judges evaluated these strategies to be higher (see, high standard deviation - SD).

n % Min Max Mean Median Mode SD
Pathos 1 25 55,5 0 95 32,1 30,00 0 22,9
Pathos 2 25 55,5 0 99 38,4 35,00 0 27,5
Pathos 3 25 55,5 0 99 30,4 24,00 0 24,1
Table 2.Descriptive statistics of Pathos evaluation scoresSource: Calculated by the author

P.1. "I really want to learn Greek, because I will feel very good." ("Θέλω πάρα πολύ να μάθω τα ελληνικά, γιατί θα νιώθω πολύ καλά." ).

P.2. "This scholarship will make my dreams come true." ("Με αυτή την υποτροφία θα πραγματοποιηθούν τα όνειρά μου.").

P.3. "I really wanted you to choose me (for the scholarship)." ("Ήθελα πάρα πολύ να διαλέξετε εμένα (για την υποτροφία).").

Pathos strategies have received the lowest suitability scores (M=33,6, collective mean for all three examples). In examples 1 and 2 above, some subjects tried to emotionally engage the evaluating body to be offered the scholarship so as to '...feel good' or make their dreams '... come true'. It can be implied that any other decision by the judges would damage their dreams, with the applicants attempting to impose an emotion of guilt on them, thus giving the judges no alternative choice (Grice, 1973) but to grant them the scholarship. In Example 3 (which received the lowest vote), 'I really wanted you to choose me', the subject is trying to secure the judges’ favourite vote with a direct emphatic statement that was found to be unsuitable.

Ethos: Regarding the ethos strategy, the teachers rated all three examples of low and medium suitability (collective mean M = 40,9), below the middle of the scale (see Table 3 for the scores for each example below). Note that the mode score is 0 (the standard deviation being high), which again shows that most judges found it totally unsuitable, while others voted higher (significant difference in the scoring between the judges).

n % Min Max Mean Median Mode SD
Ethos 1 8 17,7 0 99 44,1 43,00 0 28,3
Ethos 2 8 17,7 0 90 41,7 40,00 0 25,0
Ethos 3 8 17,7 0 95 37,0 38,00 0 24,9
Table 3.Descriptive statistics of Ethos evaluation scoresSource: Calculated by the author

E.1."Since becoming an adult, I have worked several times occasionally to contribute to family income." ("Μετά την ενηλικίωσή μου έχω εργαστεί αρκετές φορές περιστασιακά, για να συνεισφέρω στο οικογενειακό εισόδημα." ).

E.2. "I would help those who cannot speak Greek well, I would teach them a lot about life outside, how people live." ("Θα βοηθούσα αυτούς που δεν μπορούν να μιλούν καλά την ελληνική γλώσσα, θα τους μάθαινα πολλά πράγματα για την ζωή έξω, πώς οι άνθρωποι ζουν.").

E.3. "I would give them advice as if someone was bothering you, I would teach them new things like respect... towards people who are close to us even at school." ("Θα τους έδινα συμβουλές όπως αν κάποιος σε ενοχλούσε, θα τους μάθαινα καινούρια πράγματα όπως για τον σεβασμό... προς τους ανθρώπους που είναι κοντά μας ακόμη και στο σχολείο.").

The ethos in these letters is expressed in different ways. Some of those are shown in the above examples. In Example 1, the applicant attempts to project himself as a responsible family person who focuses on home over personal needs. Referring to family members is a traditional way of gaining credibility in the eyes of the receiver; which occurs typically in Eastern cultures. In examples 2 and 3, another applicant presents himself/herself as a person who lacks individualism, expressing altruism (Hofstede, 2005) by claiming to help other members of society, attempting to gain the recipient's favour. Therefore, if the applicant is chosen by the receiver, he claims that he intends to perform social work as well (a persuasive tactic found also in letters written by Germans in English). The applicants use these arguments to increase their credibility, through the use of ethos strategies, by showing that they possess social values that may be shared with the receiver. In this light, applicants undergo an identity construction process in order to present themselves as virtuous, active, and moral members of society (Psaltou & Yspilandis, 2000). Although the strategy did not gain very favourable votes from the judges, it did gain (marginally) the highest score of all the other strategies examined here. Some of the reviewers may be familiar with reading this strategy, and, in this case, they may have evaluated it more favourably, however, the mean value remained below the middle of the Ravdos scale.

Further exploration of the data

The Kolmogorov-Smirnov test was performed and revealed that the data followed a canonical distribution (p= < 0.05). In this event, the statistical tests with all the variables under consideration, being in a quantitative (scale) format due to the evaluation variables chosen for the job, were explored with the Pearson correlation coefficient. The test was used to investigate associations between suitability judgments, initially for each type of persuasive argument separately, and at a second stage between the types of persuasive strategies investigated. Statistically significant correlations (p= <0.001) were found in the judgments of the three examples of logos, pathos and ethos arguments examined separately per category. It appears that the judges evaluated analogously all three examples in each type of strategy. This reveals a high consistency per category (type of argumentation) in their judgments despite the different examples examined. In addition, statistically significant correlations were found between the different types of logos, pathos, and ethos strategies (Table 4). Notice the positive statistically significant correlations detected between all types of strategies examined and presented in Table 4.

Significance and Pearson Correlation values
Logos 1 Logos 2 Logos 3 Pathos 1 Pathos 2 Pathos 3
Pathos 1 <.001 .280 <.001 .372 <.001 .258
Pathos 2 <.001 .315 <.001 .379
Pathos 3 <.004 .219 <.001 .362 <.001 .279
Et hos 1 <.001 .239 <.001 .181 <.001 .254 <.001 .257 <.001 .267
Ethos 2 <.001 .257 <.001 .276 <.001 .356 <.001 .351 <.001 .522 <.001 .405
Ethos 3 <.001 .247 <.001 .250 <.001 .299 <.001 .271 <.001 .323 <.001 .353
Table 4.Suitability Correlations between logos, ethos and pathos argumentsSource: Calculated by the author

In more detail, those judges who voted logos strategies high, have also voted two of the pathos and all of the ethos strategies high. Those who provided a low score to logos also voted all other strategies low. This finding indicates statistically significant consistency in the voting of judges among the different types of arguments used; a judge generally votes similarly for all strategies without distinguishing between the different types. This may be an indication of a personality impact/influence and a voting behaviour (tactic). Finally, in order to examine associations with the other independent variables, cross-tabulation and correlation statistical analyses were applied, where appropriate. It was found that the judgments in these independent variables were not statistically associated with the dependent variables tested (except for the age variable, which was discussed above).

Summary and Discussion

The initial set of research questions investigated in this paper are largely addressed by the research findings. While it would be expected that the persuasive argumentative deployment of Greece and Albania was related, given the proximity of the two nations, all the strategies used were scored below the middle of the evaluation scale. Logos strategies received the highest votes, while pathos examples were granted the lowest votes. In more detail, the persuasive strategies used by these Albanian speakers of Greek were not very well accepted by the Greek native-speaker judges, who proved to be difficult to satisfy. The analysis of the data from the quantitative survey revealed the following.

1. The ethos and logos strategies received slightly higher votes (7 points higher on the scale) when compared to the pathos strategies. One possible interpretation is that in Greek culture the logos and ethos strategies are found to be more suitable than the pathos arguments. In particular, direct statements (expressed with the use of imperative), particularly those of pathos, should be avoided in letters of application for a scholarship, as they received the most negative evaluations. The use of pathos, according to Johnstone (1989), is expected to be used in Eastern societies, which is evident in the Albanian culture having been strongly influenced by the Ottoman. Notice, however, that the exact same types of argumentation tactics, expressed in a similar fashion, were used by postgraduate Greeks when writing in English for the same purpose in an earlier study by James et al. (1992).

2. The employment of ethos strategies, particularly those showing social sensitivity (examples 2 and 3), indicates that the ethical dimension in such writing seems to be shared by more cultures (Persian, Greek) and does not link solely to these Albanian authors. A characteristic paradox is documented between a) the productive use of ethos and pathos strategies by Greeks when writing in English (in the study by James et al., 1992), and b) their critical receptive assessment by Greek judges sitting in the position of the assessors in this work. This result may need deeper investigation to examine whether individuals change their reactions or attitudes towards persuasive language conventions, depending on whether they are sitting in the chair of the L2 applicant or the L1 judge (productive/receptive).

3. Ypsilandis (1994) reported that in the case of related cultures (Balkan in this case) cultures, the outcome in the use of persuasive strategies between two languages is usually successful. However, it appears that this is not the case between these two neighbouring cultures. There may be two possible interpretations: a) cultures change, and this may have occurred with the Greek case, which seems to have been adopting more Western values, gradually following a more Western type of philosophy since the 1992 study by James et al., or b) the judges in this study operated more like magisters who wanted nothing less than a perfect letter of application (particularly those of younger age, a result that supports the hypothesis of cultural change hypothesis).

4. Given that the standard deviation values were high in all tests related to the dependent variable (suitability of strategies), a significant number of judges’ votes deviated from the mean value on average (see Table 1, Table 2, Table 3). It may thus be concluded that the spread of voting is large, which in turn may also suggest a personality factor rather than being a representation of an ethnic/cultural group. When this fact is seen in relation to the high consistency of judgments (statistically significant correlations in the evaluation for each separate type of strategy), this may lead to an understanding that the personality-independent variable would need to be further tested. This psycholinguistic factor would need to be further explored with both: a) the teachers' judgment behaviours who are asked to decode the strategies used by the applicants, and b) the applicant’s argumentative use, as this variable can affect the strategic argumentative selection of the applicants.

5. As the level of education and the teaching experience of the judges did not statistically correlate with the judgment score, it can be argued that the personality hypothesis (above) can be further explored in the future together with the cultural hypothesis suggested in the literature. More personal information on the cognitive style of the applicants, which may be reflected in their personality and the choice of persuasive arguments, would assist in better understanding their argumentative tactic.

6. Age was the only variable to record an impact on the voting tactic, with older judges showing more tolerance and voting more favourably.

7. The new evaluation tool (scroll bar) used to record judges’ opinions in scale format proved to be optimal for the task, offering a detailed, more sensitive, accurate and invaluable understanding of judge reactions (all strategies are evaluated in the same consistent way), a finding revealed by the statistical correlation analysis of the vote behaviour of judges. Thereon, this evaluation tool could also be recommended as an instrument for applications of this type, in relation to specific criteria, which are often fuzzy, though they can be transformed and calculated numerically for more precise and sensitive judgments. This way the personality variable in the evaluation of applications could be minimised, particularly in cases in which important decisions, such as this, could affect the professional or personal life of others.

Regarding the limitations of the investigation, it should be noted that the sample of applicants was very small and not representative of the Albanian population, a population that would need to be further examined. However, the repetition of similar types of persuasive strategies by all subjects may be interpreted as a possible trend for the rest of the population, which would nevertheless need to be confirmed in further studies. In the second stage, the sample of judges came from different regions of Greece and despite its casual origin, it was not large enough to be considered representative of all teachers in the country. Future studies may need to address the weaknesses of the present, enriched with more applicants and more judges.


It can be concluded that the Greek and Albanian languages do not share a common pragmatic ground, at least in the receptive way these strategies are decoded, despite the similarities they exhibit in the productive way they encode persuasive messages. On the other hand, differences in the productive encoding of different types of strategies in several similar letter-writing studies have shown that the way of expressing persuasion differs from one culture to another despite the few common elements that exist Farnia et al., 2019; James et al., 1992, 1994; Papadopoulos & Ypsilandis, 2017; Psaltou-Joycey & Ypsilandis, 2000; Ypsilandis, 1994). Language and culture are two interrelated elements that influence each other, and therefore it is essential that learners are exposed not only to the structures of an L2 but also to the pragmatic perlocutionary force these and other communication features carry (Farnia et al., 2019). Although this linguistic component in question has proved once again to lead to the creation of PML failure with troublesome results for the applicant, it seems to be missing even in sites proclaiming to provide assistance in the writing of such letters (

Although the information offered in it is valid and beneficial, no consideration has been placed on the audience and the culture to which such letters are addressed, a variable that needs to be taken into account. Successful intercultural communication does not seem to be a one-way road for all cultures and would require that the sender of a message may need to consider the recipient variable and either: a) consider if they share common language pragmatics and become aware of possible differences in the pragmatics of the L2 communication code, or b) converge to the recipient's expectations and encode the linguistic product cautiously (see relevant accommodation theory, Giles, Taylor & Bourhis, 1973; Giles, 2016; Zhang & Giles, 2018), and thereon, follow a safer persuasive strata with more logos arguments included in their letters. On the other hand, it would be beneficial for both senders and receivers of messages to receive proper training on the subject; native speakers (receivers) in particular approach and decode the linguistic outcome produced by learners of their L1 with increased tolerance.

In the case where native speakers do not recognise that the origin of friction is of linguistic concern and not a personality issue, they violate a receptive maxim of quality politeness (a reverse case to Grice’s, 1999, productive maxims of politeness), in that they understand information in a false way, which in turn leads them, as a consequence, to generalise potentially erroneous views concerning the character or personality of the L2 speaker (Riley, 1989) which they have no ethical right to judge. However, the final choice and thus the responsibility always belong to the learners/users of the language, as they are the ones who encode strategically (Ypsilandis, 2023). In this light, learners should be allowed to decide how to proceed with their final strategic choices, provided that these are conscious/aware decisions/selections based on knowledge and are not grounded either on erroneous L1 transfer or faulty perceptions about the L2 language and culture. Following a knowledgeable route may improve the chances of success in communication and, also, the quality of writing in that particular genre (Farnia et al., 2019). The present study is presented as a trigger for further investigation of the issue.

Acknowledgement Statement: The authors would like to thank the reviewers for providing comments in helping this manuscript to completion.

Conflicts of interest: The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Authors' contribution statements: Both Mrs Vasilopoulou's and Pr. Ypsilandis' contributions were the conceptualization, methodology, formal analysis, investigation, writing-original draft, visualization, project administration, data curation and the writing-review and editing processes. The validation and supervision processes were under the authority of Pr. Ypsilandis.

Funding statements: As there was no external funding received for this research, the study was conducted without financial support from any funding agency or organization.

Ethical consideration statement: Since this study did not involve human or animal subjects, ethical considerations related to participant welfare, informed consent, and privacy were not applicable. However, ethical standards regarding academic integrity, transparency, and proper citation were upheld throughout the research process.

Data availability statement: All participants were informed that participation was voluntary and anonymous and there were no restrictions on data availability.

Disclaimer: The views and opinions expressed in this article are those of the author(s) and contributor(s) and do not necessarily reflect JICC's or editors' official policy or position. All liability for harm done to individuals or property as a result of any ideas, methods, instructions, or products mentioned in the content is expressly disclaimed.