1. Introduction

Impoliteness and disrespect in student e‑mails is an issue that lecturers are not unfamiliar with. To illustrate, the following anecdote has been disclosed to me by a former colleague who was then a teaching assistant in a French‑speaking Belgian university. In the context of her literary analysis class, she assessed BA1 students in romance philology with eight graded exercises consisting in an analysis of a French sonnet. For exercise 3, a student mistakenly selected another sonnet, which resulted in a zero mark for this exercise (the students had been informed of this calculation system). The lecturer nonetheless provided the student with feedback about her analysis and had a small conversation with her. In the evening of the same day, the lecturer received the following e‑mail from the student:

(1)	Bonjour,
	Je voudrais revenir sur la note que vous m’avez attribué[e] hier, pour l’exercice numéro 3. Certes, je me suis trompée de sonnet, mais mon travail ne mérite pas, à mon avis, le 0 que vous m’avez donné. Si je n’avais pas fait le travail demandé, évidem[m]ent que je l’aurais mérité. Le problème est que j’ai énormément travaillé pour ce devoir, mon erreur mériterait d’être pénalisée mais non considérée comme absence de travail. De plus je voudrais ajouter que le fait de nous donner les notes des travaux précédents juste avant un nouvel exercice noté peut, comme dans mon cas, diminuer la concentration et la confiance en soi. Je vous prie d’agréer mes salutations distinguées et espère que vous accueillerez mes remarques positivement.
	‘Hello, I would like to come back on the mark you assigned me yesterday for exercise 3. That is right, I made a mistake in my choice of sonnet, but my work doesn’t deserve, in my opinion, the null mark you gave me. If I hadn’t done the assignment requested, of course I would have deserved it. The problem is that I have worked a lot for this assignment, it would be fair to penalize my mistake but not to consider it as an absence of work. Moreover, I would like to add that giving us our marks on the previous assignments right before a new assignment can, as in my case, decrease concentration and self‑confidence. With best regards, I hope you’ll welcome my remarks positively.’

When she reported this anecdote, my colleague described the student’s behavior as clearly impolite. Interestingly, however, this e‑mail displays several linguistic expressions that are expected to achieve politeness such as, for instance, à mon avis ‘in my opinion’ and pouvoir ‘can’ that enable the student to decrease her commitment to her statements, the use of conditional form voudrais to soften her disagreement with the lecturer, and the formal closings Je vous prie d’agréer mes salutations distinguées ‘with best regards’ (I have underlined these expressions in the original message above). What may have triggered a negative evaluation of the e‑mail is the expression of disagreement towards the rules imposed by the lecturer and the overt criticism of the lecturer’s behaviour. In addition, the emphasis on the positive outcome of the student’s request to change her mark in the final sentence of the message ([J’]espère que vous accueillerez mes remarques positivement: ‘I hope you will respond positively to my remarks’) is likely to reinforce the threat entailed by the cumulated speech acts of request, criticism and disagreement, just as using the adverbial upgrader de plus ‘in addition’. Summing up, the (im)politeness assessment is based on a combination of content (the types of speech acts performed and their propositional content) and form (use of modification such as expressions mitigating and aggravating the strength of the request and criticism). Crucially, the use of verbal mitigation is not sufficient to compensate for the threats to the recipient’s positive face—by disagreeing and criticizing her, the student goes against the lecturer’s public self‑image of a person who seeks approval—and negative face—the lecturer’s freedom of action is restrained because she is requested to do something she is not necessarily willing to do (Brown & Levinson, 1987). After receiving this e‑mail, and with the support of the senior lecturer who was responsible for the course, the lecturer sent the students general information about the expression of respect in exchanges about—and the raison d’être of—the observance of instructions at the university. Even though one cannot be certain that the student who sent the e‑mail in (1) wasn’t intentionally impolite, a plausible analysis amounts to considering that the lecturer’s perceived impoliteness is caused by the student’s lack of communicative competence in the formal context of university classes.

In the same vein, it has been shown that students experience difficulties writing e‑mails in an academic context, in particular in their attempts to request favours from lecturers (e.g., Chen 2006; Félix‑Brasdefer, 2012; Lewin Jones & Mason, 2014; Savič, 2018). This is especially true when they are not native speakers of the target language (Economidou‑Kogetsidis, 2011). A plausible hypothesis is that students resort to a variety of linguistic expressions (what is sometimes called “linguistic politeness”) to make their messages sound polite to their recipients (“perceived politeness”, see Decock & Depraetere 2018 on this distinction), although, as we just saw, they are not systematically successful in this regard. Investigating their use of linguistic modification, such as mitigation and aggravation, is therefore all the more relevant in the current era of digitalization. In this paper, I focus on request mitigation achieved lexically and morpho‑syntactically, which results in the softening of the directive force of the e‑mail speech act.

This paper is structured as follows. In Section 2, I review available findings bearing on French requests in general, on the (im)politeness of French requests in particular, and I show that there is a knowledge gap concerning the modification of e‑mail requests in French. To address this shortcoming of previous studies, I present, in Section 3, a small corpus of 150 e‑mail requests performed by different students and addressed to one of their lecturers. These e‑mails have been elicited by a message I sent them, inviting them to send me an e‑mail if they wanted to know their marks in my exams. I analyze the linguistic devices included in these e‑mails in terms of mitigation and aggravation of the request head act. The results of this analysis are presented in Section 4. For instance, even though they are less common that preparatory interrogatives with pouvoir ‘can’, wish declaratives and hedged performatives are frequent request forms including lexical modification. Section 5 contains a discussion of these findings. The major implication is that students use a limited number of politeness strategies, with some degree of lexical variation, to achieve an appropriate degree of politeness when they address academics. I also address the benefits and drawbacks of the elicitation method used to collect the e‑mails. Section 6 concludes and outlines several suggestions for further research on perceived (im)politeness in e‑mail communication.

2. Previous research

Most empirical studies of the linguistic realization of requests concern cross‑cultural differences in the use of request strategies and politeness devices. An important observation is that there is a cross‑linguistic preference for conventionalized indirect expressions of the type Can you + verbal phrase (VP)? (see e.g., Flöck, 2016 for English; Le Pair, 1996, 2005 for Spanish and Dutch, respectively). However, empirical research concerning the production of requests by native speakers of French is scarce.

That being said, French requests in conversational contexts and daily interactions have received some attention in the literature. For instance, Kerbrat‑Orecchioni (2004) analyzed the requests performed by customers in a bakery. She found that imperatives were very rare, and always modified with s’il vous plait ‘please’.1 The vast majority of forms were indirect requests consisting in wish declaratives such as (2), declaratives with future tense such as (3), and availability interrogatives such as (4).

(2)	J’aimerais NP.
	‘I would like NP.’
(3)	Je prendrai/vais prendre NP.
	‘I’ll take NP.’
(4)	Vous avez NP ?
	‘Do you have NP?’

By contrast, few studies have addressed the form of requests in written French. French requests have been studied from a theoretical perspective (e.g., Manno, 2002), but, unfortunately, corpus‑based contributions are lacking.

One relevant study is Warga’s (2005), in which request forms were compared between native and non‑native speakers of French—I only report the results for the native speakers. Combining a discourse completion task (DCT) method with a spoken version of the DCT questionnaire, Warga reports that conditionals were more frequent than present tense forms. In addition, indirect requests such as Can/could you VP? were the most frequent request forms (77%), ahead of, for instance, wish declaratives (10%) and hedged performatives (9%) such as Je voudrais vous demander de corriger ma composition d’allemand (‘I would like to ask you to revise my German essay’).

Using a similar methodology, Van Mulken (1996) compared request realization in business French and Dutch. She instructed native speakers of these two languages to write a request that an employee re‑schedule his holiday because of a very busy month. Van Mulken discussed, among others, the frequency of s’il vous plait and other politeness markers in French and Dutch. In her French data, the most frequent request forms were preparatory interrogatives (61%)—53% of which were ability interrogatives—, imperatives and obligation declaratives being very rare (3% and 2%, respectively). As for internal modification, that is, modification within the request utterance itself, 80% of the requests involved the conditional, but only 4% included s’il vous plait ‘please’.

While (im)politeness and (in)directness in students’ e‑mails have been the target of recent research, for instance in English (Biesenbach‑Lucas, 2006, 2007), to the best of my knowledge, no experimental studies concerning the form of students’ e‑mails (not to mention e‑mail requests) in French have been conducted.

One exception is Ruytenbeek (2020), who investigated the politeness of Pouvez‑vous VP ? in French e‑mail requests in relation with sender‑recipient social status asymmetry. In a production task involving native speakers of French, I demonstrate that students do not use Can you VP? more often when addressing higher‑status individuals such as a Faculty Dean, in comparison with their e‑mails addressed to fellow students. Rather, they express some degree of deference towards the higher‑status Faculty Dean by using formal greetings (this was the case of 87% of the requests addressed to a higher‑status person). Concerning the relative frequency of pouvoir ‘can’ in preparatory interrogatives, in Ruytenbeek’s data preparatory interrogatives accounted for 69% of all request forms, and 55% of the requests included pouvoir in their main clause. Lexical variation within the 77 preparatory interrogatives about ability/possibility was weak, i.e., 64 included pouvoir in the second person plural of the present tense, 12 included possible (in Est‑il possible de VP? ‘Is it possible to VP?’), and only one contained en mesure de (Êtes‑vous en mesure de VP? ‘Are you in a position to VP?’). Other request forms, such as explicit performatives and hedged performatives, were quite rare in the data; imperatives and obligations declaratives were completely absent. From the perspective of internal request modification, students used hedges (dans la mesure du possible ‘whenever possible’), the politeness device s’il vous plait ‘please’, imposition minimizers such as juste ‘just’, ‘only’, and purpose clauses (e.g., pour que je puisse VP ‘so that I can VP’).

To shed light on the lexical modification of electronic requests in French, I present, in the next section, a small corpus of authentic e‑mail requests performed in an academic setting. Such an approach is welcome for at least two reasons. First, speech act modification in general, and that of requests in particular, has not yet been investigated on the basis of e‑mail data in French. Second, the variety of linguistic devices used to soften the face‑threat entailed by French requests is ill‑documented. There are, in addition, two major problems with previous empirical research on French written requests. To begin with, DCT questionnaires have been criticized on the grounds that the data collected with this method are not representative of the participants’ actual behaviors (see Flöck, 2016, pp. 43‑59; Yuan, 2001 for a discussion). This criticism is supported by substantial differences in the relative frequencies of request strategies according to whether the data are elicited or not. For instance, Flöck (2016, pp. 216‑217) shows that, in the request data collected with DCT, interrogatives questioning the preparatory conditions of requests are overrepresented, but the variety of request forms is underrepresented. In addition, in DCT studies mitigation as external modification is overrepresented, while aggravation and lexical mitigation are underrepresented. A related problem, which concerns corpus‑based approaches to request production, is that they do not control for some possible sources of variability in the data. Such uncontrolled parameters include, to only name a few, demographic and sociolinguistic differences between e‑mail senders and recipients, differences in the content of the request, or in the course of action in which the performance of the request takes place.

These methods to investigate request production are thus far from perfect. Unlike these approaches, I resort to a different methodology that consists in asking my students to send me a personal e‑mail so that I can inform them about their marks. The request data is thus “naturally elicited”, and not part of an experimental setting or the output of a DCT questionnaire. I will discuss in more details, in Section 5, how this approach, while complementing the findings of previous studies, provides additional methodological benefits.

3. Methodology

3.1. Hypothesis

In the empirical approach reported below, I elicit, and, at the same time, collect request e‑mails from students in the naturalistic context of their university studies. This is different from previous approaches, which generally involve scenarios in which a student requests the notes of a fellow student (for an overview, see Ogiermann, 2009). My approach also contrasts with previous empirical studies, in which the collected e‑mails differed in their propositional content: here, all the students will be invited to perform one and the same request.

I tested four predictions about the form of the requests that the students were expected to produce as a response to my initial e‑mail.

First, assuming that preparatory interrogatives involving pouvoir Pouvez‑vous VP? ‘Can you VP?’ are used when no specific obstacle to the recipient’s compliance with the request is identified (Francik & Clark, 1985; Gibbs, 1986), this construction is expected to be the most frequent in the data. This is because the lecturer himself informed the students that their marks were available and they would be informed after sending him an e‑mail, which makes it very unlikely that the students would identify an obstacle to receiving their mark.

Second, in line with previous empirical findings about French requests, pouvoir should outnumber other verbs and other expressions, such as possible and être en mesure de ‘be in a position to’, in preparatory interrogatives.

Third, one major reason for which students would ask the lecturer about their marks is that they are interested in knowing whether they passed the exam or not. This should reflect in their request strategies: a significant number of requests are therefore expected to consist in wish declaratives, i.e., declaratives with a verb such as to want, to wish as main verb, e.g., J’aimerais que vous me donniez ma note ‘I would like you to send me my mark’.

Fourth, following Warga’s (2005) and Ruytenbeek’s (2020) findings of a high percentage of requests with the conditional, I predict that most e‑mails will include morpho‑syntactic modification of the request by means of a verb in the conditional tense.

No specific predictions were made concerning the relative frequency of lexical items used to achieve politeness, such as s’il vous plait ‘please’.

3.2. Procedure

3.2.1. Participants

The students whose e‑mails were collected from my mailbox consisted of 150 native speakers of French, the large majority of them having the French nationality (1^st, 2^nd and 3^rd study years in language science, 143 female). They all were students at the Université de Lille SHS (Humanities) and part of the language sciences program. None of them had been informed about the research questions addressed in this paper at the time the e‑mails were collected.

3.2.2. Method of data collection

The students who took at least one exam with their lecturer (me), were told, in a general e‑mail addressed to multiple students, that they would receive their marks providing they send an individual request to the lecturer. The initial e‑mail from the lecturer was the following:

(5)

Chères étudiantes, chers étudiants,

Les notes pour le CC1 [contrôle continu] sont disponibles (nous ferons ensemble une correction lors du cours de la semaine de rentrée). Merci de m’envoyer un email individuel pour recevoir votre note (NE PAS répondre à ce message). Bien à vous, (Nom) English translation: ‘Dear students, the marks for the first exam are available (we’ll correct it together during next week’s class). Please send me an individual e‑mail to receive your mark (DO NOT reply to this message). Best wishes. (Name)’

I ensured that all the participants understood that they were supposed to perform a request by making this clear in the wording of my e‑mail (Merci de m’envoyer un email individuel pour recevoir votre note). It was therefore expected that the e‑mails sent by the participants would include a request that the lecturer inform them about their mark. Nothing prevented the participants from including additional information in their e‑mail, but such extra information is not discussed below. A handful of e‑mails were not included in the study because they were multiple e‑mails sent by students who took several of my courses.

3.2.3. Data collection and analysis

The collected data contain one e‑mail from each student. In line with recent work on e‑mail communication (Mazancourt, Couillault & Recourcé 2014), all e‑mails were anonymized by removing the students’ and the lecturer’s name (no other personal information was included in the e‑mails).2 This ensures that the present work complies with general data protection regulations (GDPR), in particular with the new GDPR 2016/679. Treatment of electronic data was done in accordance with the relevant French Commission Nationale de l’Informatique et des Libertés (CLIL) and EU legislations, and with the EU Directives 2002/58/EC and 2006/24/EC.

The analysis of the structure and the linguistic content of the e‑mails was conducted in line with Blum‑Kulka and Olshtain’s (1984) Cross Cultural Speech Act Realization Project (CCSARP) coding scheme (see also Blum‑Kulka, House & Kasper, 1989), which was complemented by small additions made by Biesenbach‑Lucas (2006, 2007) and Flöck (2016). An important difference with the original CCSARP, however, is that the classification of request forms according to “levels of (in)directness”, which is actually ambiguous between linguistic and perceived (in)directness (Decock & Depraetere, 2018), was not taken into account in my research.

In many cases, an e‑mail consists of several utterances that have different pragmatic functions. In line with the CCSARP approach, I systematically distinguish the “request head act”, that is the utterance by means of which the request is communicated, from the other utterances by means of which the students ground their request, check whether the preconditions for their request obtain, introduce themselves, etc., and which are considered peripheral to the request head act.3 These modifications, which are peripheral to the request head act, are called “external modification” in the CCSARP coding scheme. Examples of modification that is internal to the request head act, such as s’il vous plait ‘please’ and donc ‘so’, are provided in Appendix 1.

4. Results

4.1. Preferred verbs in request strategies

In line with available empirical evidence for written French, preparatory interrogatives of the ability/possibility type such as (6) were the most frequent request forms (74 out of 150 requests) (only one occurrence concerned the availability of the requested information) (cf. Table 1).4 Imperatives and obligation declaratives did not occur at all in the data.

(6)	Pouvez‑vous	me	donner	ma	note	de	CC	s’il vous plaît ?
	Can.2PP‑you	me	give.INF	my.FEM	mark	of	exam	please?
	‘Can you give me my mark for the exam please?’

 

In line with previous findings, pouvoir ‘can’ was preferred over vouloir ‘to want’ as the request’s main verb in the recipient‑oriented utterances. Alternative expressions, such as possible and être en mesure de ‘be in a position to’, were not found.

It is worth noting that the verb pouvoir, as in Pouvez‑vous VP?, can also be understood as a permission (see, e.g., Desclés, 2003). However, such an interpretation was rather unlikely in the context of this e‑mail data collection. Pouvez‑vous VP? could reasonably be taken as a question about the recipient’s permission to do some action in a situation where the issue of granting permission is relevant. But this was not the case in the present context, as the recipient (the lecturer) had already made clear he was willing to inform the students about their marks. In addition to its addressee‑oriented uses, pouvoir was also used in speaker‑oriented constructions, i.e., constructions with a verb in the first person singular, as in (7), on 25 occasions (one third of all preparatory interrogatives).

(7)	Puis‑je	avoir	ma	note	du	Contrôle continu ?
	Can.1PSG‑I	have.INF	my.FEM	mark	of the.MASC	exam?
	‘Can I have my mark for the exam?’

Like Pouvez‑vous VP?, the speaker‑oriented construction Puis‑je VP? also has a permission reading; as for the former construction, this reading was highly unlikely in the context of the student‑lecturer e‑mail exchange.

A little more than one fourth of the requests were of the wish declaratives type, and a little less than one fourth of the hedged performative type, illustrated in (8) and (9), respectively.

(8)

Suite à

votre

mail,

j’

aimerais

connaître

ma

note

concernant

le

partiel

de

pragmatique.

Following

your

e‑mail

I

like.PSG.COND

know.INF

my.FEM

mark

concerning

the.MASC

exam

of

pragmatics

‘Following your e‑mail, I would like to know my mark for the pragmatics exam’

(9)

Je

vous

envoie

ce

mail

afin de

connaître

ma

note

du

CC1.

I

you

send.1PSG.PRST

this.MASC

e‑mail

to

know.INF

my.FEM

mark

of the.MASC

first exam

‘I am sending you this e‑mail to know my mark for the first exam.’

Concerning the verbs used in wish declaratives, of the 44 requests exemplifying this strategy, aimer ‘to like’ was the most frequent verb (43,2%), ahead of souhaiter ‘to wish’ (34,1%) and vouloir ‘to want’ (20,4%) and only one example of désirer ‘to desire’ (2,3%). A plausible explanation why désirer was used only once in wish declaratives is that, in French, the noun désir and the verb désirer have a sexual connotation (cf. Franckel & Lebaud, 1990), which may have prevented students from using it to request something from their lecturer.

4.2. Internal request modification

59,3% of the requests included a form in the conditional (mostly the verb in the main clause of the request head act), which makes it the most frequent type of verbal modification achieving politeness in the data. Less than one third of the requests involved the politeness marker s’il vous plait, but other lexical modifiers, such as hedges and grounders, were rare (see Table 2).

Only one request included an exclamation mark that could be considered as aggravating the force of the request.5

4.3. External request modification

External request modification includes a variety of expressions, some of them performing speech acts on their own. In my data, salutations (98%), formal closings (60,66%), thanking the recipient in advance (57,33%), and mentioning the recipient’s title (Monsieur) (48%) were the most frequent external request modification. By contrast, less frequent external modification achieving politeness consisted in additional greetings (18,66%), self‑introduction (17,33%), and using the recipient’s name (7,33%). Examples of these expressions are provided in Appendix A.2.

5. Discussion

The findings reported in this paper largely confirm the predictions outlined in Section 3.1. First, in line with previous empirical evidence, preparatory interrogatives centered on pouvoir ‘can’ were the most frequent request forms. Second, the students displayed their interest in knowing whether they passed the exam or not, which translated into their use of wish declaratives involving aimer ‘to like’, souhaiter ‘to wish’ and vouloir ‘to want’ quasi systematically in the conditional tense; this was the second most frequent request strategy. Third, a majority (almost 60%) of all requests included a verb in the conditional, which confirms the findings of Van Mulken’s (1996) and Warga’s (2005) studies, according to which French requests are more often in the conditional than in the present tense. A detailed analysis of speech act modification in my data revealed a variety of strategies used to mitigate the threat to the recipient’s negative face entailed by the performance of the request: the speech acts of thanking, apologizing, and greeting, and the use of grounders and formal closings. In addition to his negative face, the recipient’s positive face (his want to be publicly approved of) was sometimes stimulated with the use of his title and/or his name.

In contrast to these expressions resulting in request mitigation, lexical and syntactic upgraders aggravating the force of the request were very infrequent in the data. Taken together, these results are interesting, because they show that, in spite of a context that entails a very weak degree of face threat to the positive and negative faces of the recipient lecturer, students both resorted to positive and negative politeness strategies.

Despite many requests containing internal and external modification, however, a significant number of e‑mails lacked the “polite” expressions that characterize formal communication, such as addressing the lecturer with monsieur ‘sir’ and mentioning his title and/or name, and using greetings to communicate a positive attitude towards the recipient. The results of data analysis indicate that request strategies are not extremely varied, consisting in preparatory interrogatives about ability/possibility, hedged performatives, and wish declaratives, with little variation in the choice of lexical items within a particular strategy.

Concerning the methodology I used to collect the request data, this original technique avoids the major drawbacks of discourse completion tasks often used in empirical pragmatic studies, while providing additional advantages. First, unlike DCT questionnaires, this method of data elicitation does not result in a lack of naturalness. On the contrary: the request e‑mails were part of an authentic interaction involving the students and their lecturer, which makes redundant the use of a more or less artificial description of a “speech act situation”. Second, the methodology presented in this article enables, like the DCT method, a high degree of control over the nature and the content of the speech act to be performed by the participants. The content of the request, i.e., a request for information about their mark, was identical for all participants, the relative social status between the participants and the e‑mail recipient was constant (all participants were students and there was only one recipient, i.e., the lecturer), and the general context of the data production was the same e‑mail interaction for all students. Third, the content of the speech act expected from the participants can be modified according to the researcher’s hypothesis. Such a method of naturalistic data collection can also, as DCT questionnaires, be applied to other languages, allowing for comparative pragmatic studies. Furthermore, socio‑pragmatic variables concerning e‑mail senders, such as level of studies (BA vs. MA), type of studies (e.g., Humanities vs. Exact Sciences), and concerning e‑mail recipients, such as years of teaching experience (junior vs. senior lecturer), familiarity between senders and recipients, are also amenable to manipulation.

Furthermore, additional benefits are associated with this method. It makes it possible to collect large amounts of authentic data if multiple researchers elicit data from their students using an identical e‑mail to trigger the students’ responses. This method is “cheap”, as no questionnaire administration is involved, and data collection takes place online. This method is not limited to the speech act of requesting, but it can be applied to other speech acts: assertives (provide information, answer questions, express an opinion, react to new information) as well as other subtypes of directives.

Such a method of speech act elicitation, however, is not perfect. For instance, in the present research it entailed a highly constrained context grounding the requests performed by the student participants. This may arguably have reduced lexical variation in request realization because the lecturer himself had removed possible obstacles to his sending their marks to the students beforehand. As a result, non‑conventionalized indirect request forms such as hints (I was wondering whether you had already graded our exams) and their negotiation of the requested information were extremely infrequent.

Even though this contribution to the field of politeness research is mostly empirical, my results also have theoretical implications. First, they confirm the relevance of the distinction between external modification and internal modification in speech act realization. Second, an implicit theoretical assumption I made is that some linguistic constructions are more or less conventionalized than others for the performance of some speech act, and that these differences boil down to differences in terms of relative frequency of occurrence. This relates to Terkourafi’s (2015) frame‑based approach to (im)politeness conventionality, which she defines in terms of a relationship between an expression, a speaker, and a context. Accordingly, we can consider that, with respect to the verbs désirer and souhaiter, wish declaratives including vouloir in the conditional form have a higher degree of conventionalization in the performance of requests in the context of a formal interaction following an invitation by the lecturer.

6. Conclusion

In this paper, I addressed the use of modification in e‑mail requests from French university students. I collected 150 e‑mail requests performed as part of an authentic interaction between the students and their lecturer, and analysed the lexical and morpho‑syntactic modification included in these requests. I confirmed previous studies’ findings that pouvoir is the most frequent verb in “preparatory” request forms. From a methodological perspective, I proposed an innovative method for gathering authentic speech act data, avoiding the downsides of discourse completion tasks while maximizing the number of advantages resulting from a naturally elicited data collection.

This research not only documents the lexical modification of French e‑mail requests, it has direct applications in other fields. For instance, this research is relevant to higher education and French foreign language (“FLE”) teaching specialists. Nowadays, a major societal challenge concerns adaptive difficulties for university students that arise from the variety of discursive genres they are expected to master during their studies, and the corresponding levels of politeness and formality. This highlights the importance of guidelines for formal communication at the university, and this research provides data that is amenable to such guidelines. Even though this question can only be answered on experimental grounds, it is plausible that the more downgraders and mitigation markers (and the fewer upgraders and aggravation markers) are included in an e‑mail to a lecturer, the more polite this e‑mail will be perceived by the lecturer. In addition, the interaction of request modification and speech act content on lecturers’ (im)politeness perceptions, which I illustrated by an anecdote in the introduction of this article, deserves further investigation.

Finally, analyzing request modification in terms of downgraders and upgraders, i.e., linguistic devices that contribute mitigation and aggravation to a speech act provides a means to operationalize (im)politeness in experimental studies. In particular, all other things being equal, I expect that speech acts including some degree of linguistic mitigation/aggravation should trigger psychophysiological responses corresponding to the (im)politeness assessments based on these mitigation/aggravation devices. The empirical research presented in this article thus lays the ground for experiments on (im)politeness perception in online communication.