A cross-sectional analysis of teacher-initiated verbal humor and ludic language play in an English as a foreign language (EFL) context

Abstract Although research on second language (L2) humor and language play is burgeoning, most previous studies have addressed language learners. Thus, L2 teachers in general and EFL teachers in particular have comparatively received much less attention in the literature. The present research, to my knowledge, is the first study in the Iranian context and one of the few studies in an (Asian) EFL setting that examines both qualitatively and quantitatively teacher-initiated verbal humor and ludic language play across proficiencies. In this study, within an ethnographic research design, the video-recorded talk-in-interaction of 12 Iranian EFL teachers were scrutinized to discern how they employed linguistic devices to create humorous language play in various forms and for different purposes. Additionally, the teacher participants’ self-reports of humor use as individuals and their follow-up interview responses were taken into account in further analysis and discussion of the findings. Among other things, it was found that learners’ L2 proficiency could contribute to the extent and type of teacher humor. Moreover, such humor might be employed (and “extended”) to pursue (pro)social, affective and/or managerial objectives. In this paper, considering the participants’ attitudes and concerns regarding humor use, suggestions for successful implementation of teacher humor were also put forward.


ABOUT THE AUTHORS
Mohammad Ali Heidari-Shahreza is an assistant professor of applied linguistics at Islamic Azad University, Iran where he teaches undergraduate and postgraduate students of applied linguistics. He has also worked as an EFL instructor/ teacher educator at different Iranian language institutes. He has mainly researched in the fields of second language (L2) acquisition (e.g. L2 vocabulary acquisition), gender differences and L2 pragmatics. He is currently working on different aspects of humor particularly from a pedagogical viewpoint. So far he has addressed Persian stand-up comedy from a rhetorical perspective, Persian verbal humor including jokes, para-jokes and meta-jokes from a sociolinguistic/relevancetheoretic standpoint and L2/pedagogical humor and language play from the vantage point of learner humor, teacher humor (the present research), humor-integrated language learning (HILL), etc. His recent studies have been published/ accepted in the Language Learning Journal (Taylor & Francis), HUMOR Journal (De Gruyter), Gender Issues (Springer) and International Journal of Applied Linguistics (Wiley).

PUBLIC INTEREST STATEMENT
Humor still seems a controversial issue in (language) education: On one hand, some teachers frown upon humor as a threat to their class authority and the serious business of teaching. One the other hand, some others based on their own experience advocate (a sense of) humor as an instructional tool with far-reaching benefits. This article explores different aspects of teacher humor. It begins with defining humor and highlighting its pedagogical benefits. Then, natural instances of language teachers' humor are analyzed to see how they may function within an educational setting. This study opens a window on teacher-learner humorous interaction at different levels of language proficiency. It addresses teachers' concerns in using humor and suggests practical guidelines for successful implementation of humor. The present research can particularly be insightful to those who are looking for effective ways to optimize learning and teaching experience.

Introduction
The present research primarily focuses on teacher-initiated (verbal) humor and (ludic) language play. Thus, it seems illuminating to initially have a brief look at their definitions and how these two notions may converge and/or diverge. A glance at their relationship is particularly important as the relevant literature generally seems less inclusive of and sensitive to the two terms, ignoring or preferring one over the other, if not treating them both indiscriminately (see also Bell & Pomerantz, 2016).

Humor and language play: Definitions and relationship
Despite different approaches, humor is generally defined as the creation and communication of (conceptual) incongruity, often resulting in its audience's mirth (Banas, Dunbar, Rodriguez, & Liu, 2011). Humor, based on four overarching notions of "language", "logic", "identity" and "action", can respectively be classified into four categories of "verbal", "ideational", "existential" and "physical" humor (Heidari-Shahreza, 2017;Juckel, Bellman, & Varan, 2016). Verbal humor, thus, refers to the language-based creation of a humorous message employing linguistic devices such as pun, allusion, irony, etc. Similarly, language play is a metacognitive activity that entails (creative) alteration of linguistic elements, available in a given context, to create humor and/or to serve other (e.g. language learning) objectives (Bell, Skalicky, & Salsbury, 2014;Cook, 2000). Generally, there are two prominent approaches to language play in the existing literature of second language acquisition (SLA): (a) "ludic" language play 1 that is better known since Cook (1997Cook ( , 2000 and mainly spins around the notion of language play as (self-)amusement (Broner & Tarone, 2001) and (b) language play "for rehearsal" which was put forth by Lantolf (1997) on the grounds of Vygotsky's "private speech" and regards language play primarily as a means for language practice and development.
Therefore, it can be said that verbal humor is most akin (and related) to language play for fun (see Figure 1). Pointing to the same relationship, Forman (2011) considers (language) play as a form of "verbal art" and "linguistic creativity" which might also turn to be humorous. This paper, as far as its methodological underpinnings and space considerations permit, mainly addresses both verbal humor and ludic language play. Also, it tangentially considers physical humor for, logically speaking, classroom humor may have an action-based outlet as well. Nevertheless, for the sake of brevity and coherence, henceforth, the word "humor" is used as an umbrella term for these notions unless a distinction is due. To pave the way for the present research, in the following sections, first, the use and potential benefits of teacher humor in classroom setting (particularly within an EFL context) are elaborated on. Then, the relevant research and rationale for this study are briefly addressed.

Teacher humor: Functions and benefits
While developing pragmatic competence is a part and parcel of SLA, EFL context often falls short of providing the learners with a rich, authentic input to notice and possibly acquire the forms, norms and constraints of L2 pragmatic use. Humor, in this regard, can be an effective instrument to teach sociocultural and pragmatic conventions of a second/foreign language, compensating for the lack of subtleties of language use in the impoverished EFL input (Washburn, 2001). As Deneire (1995, p. 295) points out "well-developed communicative competence implies humor competence, and vice versa". Thus, humor can highlight L2 sociopragmatic particularities and add greatly to the learners' cultural awareness.
Furthermore, Dörnyei (2001) includes humor in the list of the most motivating features of language task content. He also recommends teachers to "bring in and encourage humor" as a potent but neglected factor to lighten classroom atmosphere (p. 42). From the perspective of "emotional response" theory (Mottet, Frymier, & Bebee, 2006), positive affect as induced by appropriate teacher humor encourages more "approach" and fewer "avoidance" behaviors which, in the learning environment, can be conducive to more positive attitudes, higher motivation and better academic performance on the part of the learners. Likewise, humor may serve as an attention-raising and input-enhancing tool at different levels of language. Deneire (1995) sets specific examples for the use of humor to draw learners' attention to linguistic subtleties at phonological, morphological, lexical and syntactic levels which otherwise could be demanding for teachers and unengaging for learners to deal with. Here, is an example at syntactic level: Student 1: "The dean announced that he is going to stop drinking on campus." Student 2: "No kidding! Next thing you know he'll want us to stop drinking too." (p. 290) But, probably the most significant function of humor in educational settings is its potential to enhance learning (Wanzer, Frymier, & Irwin, 2010). Despite mixed findings in the literature (see Banas et al., 2011 for a review), it seems, student learning and retention can be improved if appropriate, relevant humor is persistently incorporated at the right points of instruction. As Garner (2006, p. 180) asserts "[t]he 'ha-ha' of humor in the classroom may indeed contribute to the 'aha!' of learning from the student."

Research on teacher humor
Many language teachers, based on their teaching experience and intuition, have come to realize the beneficial effects of employing humor (Bell, 2009(Bell, , 2011. Nevertheless, humor in the language classroom (or classroom humor) still seems to have a long way to go. This is particularly the case in formal (, traditional) contexts of language education where teachers may need to be more cautious about what is contextually/culturally permissible to be included in their teaching practice. Some teachers in such settings might be concerned about how their humor may be perceived and judged by others in and outside of the classroom (e.g. students, their parents, school board). They may also regard humor as a harmful or (at least) a trivial epiphenomenon of class interaction and thus decide not to invest on it (see also Davies, 2015).
Such concerns (and controversies) regarding humor use in language education call for further research. In fact, despite recognizing its significance, even being regarded as the fifth component of communicative competence long ago (Vega, 1990), "the study of the use and understanding of L2 humor has been largely neglected within this paradigm [L2 pragmatics]" (Bell, 2011, p. 1). Moreover, the scholarship on L2 humor has mainly focused on language learners and only few studies have addressed the role of teachers (Sterling & Loewen, 2015). Forman (2011, p. 544) highlights the significant role of language teacher as the one who "calls the shots in a whole-class setting" and decides upon "what is admissible as play -its nature and the degree to which it may occur". Nevertheless, he states "there is little enquiry to date into the role of the teacher in producing and enabling language play in L2 classrooms" (p. 559). Likewise, Banas et al. (2011), in a review of four decades of research on instructional humor point out "[s]tudies that examine instructors' self-reports of humor use have provided a starting point for research on humor in the classroom, but there is a dire need for more naturalistic research, both descriptive and experimental" (p. 137). Thus, although it seems research on L2 humor, inside and outside classroom setting, is burgeoning, teachers' use and perceptions of humor, especially in the Asian EFL context, await further research (Petraki & Pham Nguyen, 2016). In the following lines, the few studies with a special focus on the use of humor by (foreign) language teachers are briefly reviewed to pave the way for the present research (see Banas et al., 2011 for related studies in other educational settings and Bell, 2011 for humor scholarship in teaching English to speakers of other languages (TESOL)).
Sullivan (2000a) examined a Vietnamese EFL teacher's class discourse and performance with a focus on storytelling and wordplay. The analysis of her two-month observation of two Vietnamese, university-level classes (taught by the same teacher) revealed that the spontaneous instances of storytelling and wordplay (e.g. repetition, pun, synonymy) created a joking atmosphere which could contribute to the learners' intrinsic motivation, further engagement in language use, Vygotskyan expert-novice interaction, and authentic focus on form (see also Sullivan, 2000b). Similarly, Forman (2011) explored language play and verbal humor instigated by a bilingual instructor in a post-beginner class of Thai EFL learners. The textual analysis of the audio-recorded classroom data revealed that the teacher's humor mainly took linguistic and discursive foci and could have affective, sociocultural and linguistic benefits for the learners. Being competent (enough) L2 users and having control over classroom activities, Forman concludes, language teachers are able to instigate humorous language play and draw learners' attention to various aspects of language development, "while also avoiding the sterile way in which such a focus has been achieved in traditional syllabuses" (Cook, 2000, p. 193). Sterling and Loewen (2015) also analyzed instances of teacher-initiated language play and playful language-related episodes (LREs) in an intermediate class of Spanish as a foreign language. Analyzing audio-and video-recorded data of the whole-class interaction, they reported that despite relative pervasiveness of LREs and to some extent language play, "playful LREs accounted for 4.6% of all LREs, 13% of all language play episodes, and 0.5% of total class time" (p. 78). Furthermore, vocabulary, compared to grammar or pronunciation, was found to be the most frequent target of language play. In a recent, ethnographic study, Petraki and Pham Nguyen (2016) focused on teacher humor in a Vietnamese EFL context. The analysis of video-recorded instruction and field notes revealed that spontaneous humor especially in the form of "humorous comments", "jokes" and "funny stories" were most preferred by the 30 tertiary-level teacher participants. Follow-up interviews also indicated that the teachers generally supported the use of relevant, contextually appropriate humor to enliven the classroom atmosphere and improve teachers' immediacy. However, the participants who did not use humor (during observation sessions), were concerned about the nature of the courses they taught (i.e. the topic), learners' serious personalities and class activities (e.g. their presentations).
This study hopefully contributes to the scant literature on (EFL) teacher humor by examining the classroom humor and language play of 12 male and female Iranian EFL teachers across elementary, intermediate and advanced proficiencies. To my knowledge, this is the first study in the Iranian context and among the few ones in the (Asian) EFL settings with an overriding focus on teacher humor from both a qualitative and quantitative perspective. As the reviewed studies above imply, also, it is probably the first study to compare and contrast the use of teacher humor across different levels of L2 competence. As Petraki and Pham Nguyen (2016, p. 99) point out "[a]n EFL teacher employing humor must consider not only its relevance or appropriateness, but also whether it suits their students' level of English proficiency so that they can enjoy it. This "double requirement" of EFL humor promises worthwhile investigation". It is hoped that the present research aids the L2 humor scholarship in providing language teachers with practical guidelines to fully implement humor in their instruction. After all, as Schmitz (2002, p. 96) highlights: There is, without any doubt, a need for research on the use of humor in language classrooms, but until there are sufficient studies based on experiments with humor in different teaching situations, with different levels of proficiency, different target and source languages, in different countries, most of the proposals and recommendations will perforce be based on practical experience with humor and classroom teaching.

Research questions
Through its ethnographic, unobtrusive design, the present research mainly sought to find answers to the following questions: (1) What (verbal) humor techniques and forms did the Iranian EFL teachers mainly employ in their elementary, intermediate and advanced classes?
(2) To what extent did instructional specifications (e.g. ±relevance, ±L1 use) characterize (or accompany) the teacher participants' use of humor in these classes?
(3) Was there any (significant) relationship between the participants' preferred humor styles as individuals and their classroom humor as teachers?
(4) What were the participants' attitudes and concerns regarding implementing humor in an (Iranian) EFL context?

Teacher participants
On the whole, 12 Iranian EFL teachers from two language institutes in Iran took part in this research. Via convenience sampling, one male and one female teacher from each institute, teaching at one of the targeted proficiency levels (i.e. elementary, intermediate, or advanced), were selected. Thus, four teachers in each proficiency group were investigated. As can be seen in Table 1, on average, the participants were young, initial or mid-career teachers, holding mostly BA and/or MA certificates in teaching English as a foreign language (TEFL). Through reviewing their career records and initial interviews with them, it was assured that the teachers generally had the same professional, sociocultural and linguistic background. Also, the slight increase of age, experience and education with the proficiency levels they taught (e.g. 27, 29, 32 years old) did not seem to be significant, considering their overall qualifications. Likewise, almost the same teaching and administrative rubrics were observed in both institutes.

Initial interviews
At the beginning of the research, semi-structured interviews with the teacher participants were conducted (a) to break the ice and gain more insights into their teaching philosophy and practice and (b) to bridge the information gaps in their teaching records and (re)assure their suitability for the research objectives.

Follow-up interviews
To delve deeper into teacher-initiated instances of humor, another round of semi-structured, but this time video-cued, recall interviews were carried out. Although the interview questions could vary for each participant and each instance of humor, they mainly spun around "the how and the why" of using humor as well as the pedagogical effects it could have in the short and/or long run.

Humor styles questionnaire
The humor styles questionnaire (HSQ) developed by Martin, Puhlik-Doris, Larsen, Gray, and Weir (2003) was also employed to shed light on the participants' preferred styles (or types) of humor as individuals. The HSQ is a widely used, self-report questionnaire which distinguishes two styles of "positive" humor (i.e. "affiliative" and "self-enhancing") and two styles of "negative" humor (i.e. "aggressive" and "self-defeating"). It contains 32 items (eight questions for each humor style) on a seven-point Likert scale ranging from "totally disagree" to "totally agree".

Context and data collection procedure
The institutes, in which the participants taught, ran five 17-session semesters in a year. The learners attended two sessions a week during fall, winter and spring semesters and three sessions a week during the two-semester (intensive) summer classes. Out of the17 sessions in a semester, two sessions were devoted to the mid and final tests. Thus, in practice, 15 sessions were left for the actual teaching (and observation). It was decided that the fourth, eighth and twelfth sessions of each class in the fall semester of 2016 be observed because firstly the institutes could maximally allow three observations per class and secondly three rounds of observation at roughly the beginning, middle and end of a semester could reflect what was generally practiced in the classes. The learners in the observed classes did not notably differ in terms of number, gender, age, etc. On average, the classes contained 14 male and female Iranian EFL learners, mostly at mid-or late-teens. The procedure of data collection was carried out in three steps: (a) At first, necessary permissions and signed consent forms were obtained from the respective authorities and the teacher participants. Also, the initial interviews were conducted and audio-recorded.
(b) In the second phase, using a lapel microphone and a small camcorder, the 12 classes (and 36, 90 min sessions) were audio-and video-recorded by the researcher and an assistant. In all, about 52 h of instruction were observed. Following Petraki and Pham Nguyen (2016), a practice observation session (without recording) was initially run in each class to mitigate the "observer effect" and possibly neutralize the presence of a camera and a lapel microphone for both the teachers and the learners. While "no effect" cannot be claimed, in the actual observation sessions, neither learners nor teachers seemed significantly over-or underacting in the classes.
(c) After the sessions were observed, the teachers were kindly asked to answer the HSQ (see the previous section). Then, they were individually invited for follow-up, semi-structured interviews during which instances of their humor were played and discussed. As in the initial interviews, the teachers could optionally answer the questions in English or Persian. Roughly speaking, two-third of the interviews was conducted in Persian, the language in which they felt more comfortable especially when it came to critical, multifaceted viewpoints. The interviews were also audio-recorded except for two teachers who preferred their answers not to be audio-recorded. In these cases, together with an assistant, only notes were taken. https://doi.org/10.1080/2331186X.2018.1430474

Data analysis
The audio-and video-recorded data capturing teachers' talk-in-interaction, the interviews, as well as the field notes were fed into Nvivo software and analyzed at several levels: (a) In the initial analysis, tracing contextual cues such as laughter (or smile), word order, marked vocabulary, prosody (and also based on the coders' intuition), instances of teacher humor were discerned separately by the researcher and an assistant, fully cognizant of the research objectives. Several joint meetings were, then, held to discuss the humor instances and resolve points of disagreement. Few cases on which no agreement could be reached were discarded.
(b) In the second stage, different techniques of verbal humor (or linguistic devices used to create humor) such as pun, irony, or allusion were identified. In so doing, a recently modified version of Berger's taxonomy (Berger, 1976), a long-established, widely used taxonomy of humor, was employed as the basis of the identification (see Heidari-Shahreza, 2017; Juckel et al., 2016 for further information). Also, through a meta-analysis of research on pedagogical humor and consulting related reviews (see e.g. Banas et al., 2011;Wagner & Urios-Aparisi, 2011), humor forms and the instructional specifications (most) germane to the use of humor in classroom setting were decided upon and discerned in the observational data. Then, using descriptive and inferential statistics, all identified instances of humor were analyzed. Cohen's kappa was calculated as high (0.80).
(c) In the third phase, the participants' responses to the HSQ were examined to discern (significant) differences in the teachers' humor styles as individuals. Likewise, through content analysis of the follow-up interviews, major themes of humor use (especially potential benefits and concerns from the participants' standpoint) were extracted.
(d) In the final phase of the analysis, I looked at how teacher humor was initiated and developed within class interaction. Table 2 presents a micro-level account of verbal humor techniques across proficiencies (and genders). Based on the frequencies of occurrence, "pun", "allusion" and "irony" are the three most frequently employed humor techniques by the teacher participants. Furthermore, although the discrepancies in the frequency of (verbal) humor use by the elementary, intermediate and advanced teachers did not reach statistical significance (P(χ 2 > 1.44) = 0.99), the number of humor techniques seems to go up with an increase in the recipients' proficiency level; while the number of verbal humor techniques instigated by elementary teachers amounts to 11, the totals almost double and triple for their intermediate and advanced counterparts (19 and 29 respectively). It is also worth noting that a Poisson regression was run on the count data of individual humor techniques to see whether or not their increase or decrease was statistically significant (see also Cameron & Trivedi, 1998 for further justification and information on the analysis of count data). As seen in Table 2, none reached the significance level at p < 0.05.

Teacher humor forms
As can be seen in Table 3, "funny comment", "visual humor", and "physical humor" are the three most common forms of humor for the elementary teachers. This trend, however, does not persist in the higher levels: The frequency of visual and physical humor decline but more incidences of verbal (or language-based) humor such as "funny story", "joke" and "riddle" emerge. "Funny comment" also keeps a consistent, increasing pattern, topping the humor forms at all three levels. Despite such (notable) rise and fall across proficiencies, the overall variation in the humor forms employed by the teachers was not statistically significant (P(χ 2 > 10.15) = 0.42). Likewise, based on Poisson regression results, only the increase in the number of funny stories was statistically significant at p < 0.05.

Pedagogical aspects of teacher humor
From a pedagogical perspective, (at least) two general patterns of humor use are discernable in the teaching practice of the elementary, intermediate and advanced teachers under investigation (see Table 4). Firstly, at each level of proficiency, the use of "relevant", "spontaneous", "other(s)-directed" humor generally seems to outweigh its "irrelevant", "planned", "self-directed" counterpart. Secondly, higher rates of such humor seem to tally with teaching at higher levels of L2 competence. It is also worth noting that the teachers' use of L1 (i.e. Persian) to deliver humor drops as they enter classes containing more proficient EFL learners. Notwithstanding, the observed instructional differences in teacher humor across proficiencies did not hold to be statistically significant (P(χ 2 > 6.10) = 0.41). Likewise, at the individual micro-level, except for the increase in the use of relevant humor, other instructional features did not change significantly across proficiencies.

Teacher humor: Benefits and concerns
As mentioned before, via content analysis of the participants' interview responses, the major benefits of implementing humor in an Iranian EFL context (from their perspective) and the concerns they had in so doing were deduced (see Figure 3). The positive effects may broadly be classified into three categories: (a) social and affective (e.g. improving teacher's immediacy behaviors, enhancing learners' motivation), (b) pedagogical (e.g. facilitating teaching delivery, serving as a gap filler), and to a lesser extent (c) managerial (e.g. facilitating class management, mitigating teacher's face-threatening acts (FTAs)). 2 The teachers' concerns also seem to be mainly managerial, pedagogical and cultural in nature: Some of the participants saw it likely that using humor in class might function to the detriment of the class/teacher "formality", 3 "seriousness", or "discipline". Alternatively, teacher humor, in their view, could "sidetrack" the class from learning and speaking English (or being regarded as such by the respective authorities). Likewise, it was a concern how "Western" humor could be presented within an (Iranian-)Islamic context, ensuring its age, gender, and language level appropriacy. For some teachers, L2 humor could potentially bring about misunderstanding, confusion or offense because "language, culture and humor, all at the same time, could be too much a burden on the learners' minds" or the humor (e.g. jokes) could turn out to be somehow risqué and "agitate class atmosphere", "especially when both boys and girls were in class".

Teacher humor: A micro-level, interactional view
In this section, several instances of teacher-initiated humor are examined (across proficiencies) to gain more insights into how humor may be born, embedded or developed in the context of a teacher's class interaction/communication: In this incidence of spontaneous humor, instead of merely "recasting" the learner's erroneous utterance (e.g. "you make chicken sandwich at home"), the teacher takes advantage of this slip of the tongue and instigates a phonological wordplay on the words "chicken" and "kitchen". This sound play and the accompanying laughs, kill two birds with one stone by both "refreshing the class" (after the time spent on the reading) and "drawing their attention to a mistakable pair" (i.e. "kitchen/ chicken"). The next two instances of humor take into account how wordplay is employed (and becomes more sophisticated) at higher levels of L2 competence. The first one takes place at an intermediate class when the teacher is talking (partly in English, partly in Persian) about his students in the previous semester: Example 2: 01 T: I had many good students too. but a few of them were really "zarif". (Persian word meaning delicate/elegant) (1) sorry! HHH "zaeif".[(Persian word meaning weak) 02 Ss: [HHH] 03 T: ☺of course, they[were zarif too. HHH but here I meant zaeif.☺ 04 Ss: [HHH… Again a slip of the tongue forms the basis of a (phonological) wordplay (i.e. 'zarif/zaeif'). The real humor, however, lies in the connotative meaning of the word 'zarif', referring to feminine beauty. The teacher, in this natural incidence of humor, "let the students laugh at [his] blunder", "presupposing an informal relationship with the students" within the dominantly formal context of the class. Thus, unlike the form-focused, relevant wordplay in the previous example, here this off-task humor occurs when the teacher is bridging a gap between his lesson plans. It serves the purpose of "getting closer to the students" by laughing and let others laugh at a notion (i.e. there are beautiful girls in language classes) that, normally and without the pretext of humor, is not appropriate to pinpoint. Therefore, among other things, humor functions as an immediacy behavior to increase the warmth among class members and to express "I'm one of you" by the teacher. In the third example, the advanced teacher is responding to the students who insistently but indirectly ask her to cancel the planned quizzes for the coming weeks.
Example 3: (At the beginning of the class, before that day's quiz) 01 Ss: teacher! teacher! this is the last quiz, right? 02 Ss: (playfully pretending to be true) yes, this is the last quiz. (1) no more quizzes (1) please! 03 T: the last quiz?!(2) (slightly smiling) yes, this is the LAST one.
(At the end of the class) 05 T: so don't forget to study for the next week's quiz! 06 Ss: teacher! (1) but you said this is the last quiz! 07 T: ☺yes, I did.☺ HHH today's quiz was the last quiz in Azar. (Persian month, December) ☺next week, we enter a new month.☺ 08 Ss: no!☺the teacher was joking!☺ In this instance, the teacher teases the learners by crafting a semantic, context-based wordplay. The word "last" conceptually and contextually can be rendered within both a month span (what the teacher intends) and the whole semester (what the learners (wish to) "semiotize"). Unlike previous examples, the humor here consists of two detached episodes: The first one, at the beginning of the class, which is delivered in a normal, serious tone, seemingly void of any humor, and the second part, at the end of the class, especially line seven, when in terms of Attardo and Raskin's (1991) General Theory of Verbal Humor (GTVH), the "punch line" is reached, the "script opposition" is resolved, and thus the "language" component (i.e. the semantic play on "last") fulfills its function as a "knowledge resource" to create humor. That is, the learners understand the teasing intention of the teacher about what she meant by the word "last". This teasing or (aggressive) others-directed type of humor serves (at least) two purposes: In the first episode, it aids the teacher in class management by "calming down those learners who seem tired of the quizzes". Secondly, it mitigates the teacher's FTA, that is, the confrontation with some learners (particularly, those calmed down in the first episode) when she reveals her real intention to give more quizzes (as planned). Therefore, her insistence to follow the syllabus that may "taste bitter" to some learners is moderated when, thanks to the teacher's teasing, almost the whole class laughs off the "last quiz" request.

Allusion and "extended" humor
The following instances, in addition to other aspects of teacher humor especially the humor technique, "allusion", elaborate on what I would like to call "extended" humor, that is, humor extended from its original, (natural) context of occurrence and use to other similar situations. The first example is taken from the elementary level where the class is talking about the (official) holidays in the near future: Here the teacher humorously comments on the coming holidays by alluding to Delpazir (literary, pleasant), a popular, food processing company and its weekly cash gifts to randomly selected winners. This customer appreciation idea was widely advertised on Iranian TV channels as "Delpazir cash Wednesdays" roughly at the time of data collection. The teacher, employing allusion as a humor technique, joins the learners in rejoicing these red-letter days in the calendar. According to the teacher's follow-up interview, she extends the use of this humor to other classes where the learners point to the same off-school days. Thus, being successfully implemented once, this incidence of spontaneous humor, becomes a part of the teacher's "humor repertoire" to "do humor" in similar situations. This so-called metamorphosis (from a natural incidence of humor to a prepared, planned one) was observed in intermediate and advanced classes too. In the example below, one of the intermediate learners is complaining about the length (and difficulty) of the coursebook: Example 5: 01 S 1 : teacher! (resentfully) (.) we should read all /these/ pages? 02 T: yes, of course! 03 S 1 : no:! (1) it's very much, teacher! 04 T: come on! HHH (in Persian) you're not going to break the elephant's horn, are you? 05 S 1 , Ss: HHH… Without the follow-up interview with the teacher, this instance of humor particularly its "extended" feature could not be unmasked. According to him, in the previous session (not included in the observation schedule), the same learner translates the idiom "to take the bull by the horns" into Persian as "to break the elephant's horn". 4 This humorous slip that makes the whole class laugh (at the mistake), is humorously employed in the next session by the teacher to reply the same learner's complaining about reading the whole coursebook. Therefore, drawing upon allusion and intertextuality, the teacher creates in Attardo's (1994) words, the "para-humor" or the second generation of this humor and extends its use to other applicable situations as long as, according to him, his learners "remember the original humor incidence by the learner" and do not regard it as the teacher's "new, original" blunder. The following instance is recorded in one of the advanced EFL classes where the teacher is talking about different ways of improving language skills. Example 6: 01 S 1 : how can we improve our language more? 02 T: in different ways (2) listen to the English news (.) talk with your friends in English (1) even outside the class. you know,☺ there are ways to improve your English as many as there are students.☺ for example, (pointing to one of the learners) there's a way for you (.) another for the one behind you.☺ 03 Ss: (some realizing the allusion) ☺ye:s! ☺ 04 T: HHH you remember Marmoolak?! (the movie, Lizard) The teacher's humor, once again, is built upon an allusion; this time to a memorable scene from a popular Iranian movie (i.e. Lizard) in which its leading actor plays the role of a criminal in the guise of a clergyman, delivering a sermon on different ways to reach God. The teacher using almost the same words and body language (as the movie protagonist) replies his learner's question about improving language skills. Teacher's humor here is an instance of "heteroglossia" (van Dam, 2002) or Bakhtinian "double-voicing" (Bakhtin, 1929(Bakhtin, /1984 in that he shifts from his "teacher" voice into the voice of a humorous "movie character" to create his intended humor (see also Broner & Tarone, 2001). It also features "extendedness", because based on the follow-up interviews, he quotes or mocks the same or other characters when applicable to class settings. In fact, talking about cheating on finals, he used the same humor in the third observation session of his class: "students find different ways to cheat on a test (1)☺actually (.) there are ways as many as those who take a test☺".

Discussion and conclusion
In this section, as far as space considerations permit, the findings of the study are discussed and, drawing upon the relevant literature, some of the implications for the use of humor in (foreign) language settings are touched upon: Pun, allusion and irony were the most frequently used, verbal humor techniques at all three levels of language proficiency. As a general trend, more instances of such humor were employed at higher levels of L2 competence (see also Cameron and Trivedi (1998) for the statistical basis of count data interpretation). It would also probably be safe to regard the humor recipients' L2 proficiency as a contributing factor in the type and the extent of (linguistic) devices the teachers used to create humor. As Deneire (1995) notes some language teachers might assume that elementary learners fail to decode humorous communication especially when it contains ambiguous meanings (e.g. irony, allusion, joke) and thus, they may postpone the use of humor to higher levels of language proficiency. As a matter of fact, learners' low proficiency was a concern for several teachers in their humor use. In this regard, the relevant literature also suggests that more competent L2 learners may generally feel more comfortable with classroom humor (Neff & Rucynski, 2017).
Furthermore, although visual and physical humor were commonly used at the elementary level, these humor forms were replaced with more (linguistically) complex forms of humor such as funny story, joke, and riddle at higher levels (see also Petraki & Pham Nguyen, 2016). In fact, the increase in the use of funny story across proficiencies reached statistical significance. This inclination toward more language-based humor may again, among other things, point to the teachers' considering their audience as more linguistically and "symbolically" (see Kramsch, 2006) competent in appreciating their humor. Additionally, as Bell (2011, p. 12) reasons narratives such as funny stories, (some) jokes and riddles, let L2 users "obtain and hold the floor" and do not entail quick responses. Hence, they are easier to deal with and favored over other types of humor.
From a pedagogical perspective, increasingly more instances of relevant, spontaneous, other(s)directed humor and a lesser extent of L1 use were observed in classes with more proficient learners. Considering the symbiotic, co-constructed nature of (teacher) humor, as also notable in humor examples presented in the previous section, it might be said that the high-level language classes enjoyed more Bakhtinian "symbolic freedom" or the ability to creatively maneuver beyond the norms and centripetal forces of language (see Belz, 2002). Hence, teacher humor could be realized in a form which was broader in (target) coverage, more creative, and of course, less dependent on the L1. Nevertheless, other factors should not be ignored. For instance, at higher levels, more emphasis may be on L2 use or, as also implied by several participants, teachers are more likely to have a tight schedule to reach the course objectives. Additionally, more familiarity may exist between the teacher and the learners. In this regard, Bell et al. (2014) noted that the familiarity with the context, the familiarity with and the presence of equal or higher status interlocutors, learners' language-related experience and L2 proficiency contributed to the variation observed in the L2 users' language play.
In addition, the teacher participants were not significantly different in their preferred humor styles; they generally preferred affiliative and self-enhancing styles (i.e. positive humor) over aggressive and self-defeating ones (i.e. negative humor). In addition to other implications, this may indicate that the observed differences among elementary, intermediate, and advanced teachers in their classroom humor are probably not a function of the discrepancies in their humor use as individuals and contextual factors (e.g. the learners' proficiency) might have played a role as predicted earlier. Also, this study found that the participants as individuals (and not necessarily teachers) significantly preferred self-directed humor less than other(s)-directed humor. Although, a thorough treatment of this finding definitely goes beyond the present research, besides psychological factors, one may trace its roots within the framework of "face" theory and Persian culture-specific schemata. As a reviewer notes, this preference for less self-directed humor may also be tied to their occupational values and concerns (e.g. their social image as a teacher).
The participants also pointed out some of the potential benefits of employing humor such as enlivening class atmosphere, improving teacher's immediacy behaviors and facilitating instruction which are in line with what other studies have reported (see e.g. Forman, 2011;Petraki & Pham Nguyen, 2016). They, however, seemed concerned about their authority. In this regard, Cook (2000, p. 204) asserts that "[p]lay...does not entail a rejection of order or authority, though it does at least imply more voluntary and creative reasons for embracing them". Thus, although historically (and/or traditionally), some teachers may regard the use of humor as a threat to their class authority and status, appropriate humor can, in fact, enhance learners' perception of their teachers' credibility and competence (Wanzer, Frymier, Wojtaszczyk, & Smith, 2006). Needless to say, too much humor whatever may connote a clownish behavior which is to the detriment of a teacher's stature.
Some teachers were also worried that their humor might be (or be regarded as) an off-task, course-irrelevant behavior. In this regard, as Pomerantz and Bell (2011) highlight, instead of discarding or frowning upon the humorous moments in language contexts as wasteful distractions from the teaching and learning objectives, humor should be viewed as an avenue to the learners' L2 awareness and attainment. Thus, we need, in Bushnell's (2008) words, "to re-conceptualize LP [language play] as a possible motivator and facilitator [all emphasis original] rather than as disruptive, "offtask" behavior" (p. 64). As for how teacher humor might be regarded by higher status authorities, it seems in the formal (Islamic) context of education in Iran, despite promising improvements, the use of humor is not fully welcomed yet (Rashidi, Eslami, Rakhshandehroo, & Izadpanah, 2014). Thus, implementing humor in this (or any similar) context would not be without challenge (see the recommendations below).
Somehow related to this concept was the participants' concern about creating culturally appropriate and understandable humor. In this regard, it is worth noting that humor (especially the culturebased one) is, in essence, a double-edged sword and one should not expect any use of humor would result in positive effect. Notwithstanding, it should also be heeded that cultural differences may successfully be used to do (instructional) humor (Bell, 2011). The teachers, however, should not ignore the complexity embedded in (certain types of) humor as it may impede learners' perception of L2 humor without scaffolding by their teachers (Bell, 2009).
The examples of teacher humor, qualitatively analyzed in the previous section, are also insightful in different ways. In addition to what is already addressed, it is also noteworthy that teacher humor, in many cases, is born and crystallized in the context of classroom interaction, enjoying the "spoken artistry" (Sullivan, 2000a) of other interlocutors especially the learners. As pinpointed earlier, this contextually co-constructed humor can be even an extension of what has been initiated before (i.e. extended humor) by the teacher or the learners. This, in turn, calls for a broader scope of teacher humor. van Dam (2002, p. 259), in this regard, asserts "[i]f singly-authored, context-free, monolithic turns-at-talk are considered the building-blocks of "interaction", the framing practices, linguistic play [emphasis added] and emergent face work that occur in the course of discourse-units-in-progress remain invisible". Furthermore, as examples one and two, analyzed earlier, imply natural classroom humor may be triggered by learners' and teacher's mistakes and errors (or responses to them). Thus, error (correction) can be a potential source/context of pedagogical humor. That is, teachers can take advantage of such linguistic problems to instigate humor for further rapport and/ or language support (see also van Dam, 2002).
In spite of the intricate, multifaceted nature of humor and sometimes equivocal findings of humor research, several suggestions regarding the use of humor in classroom can be made: Firstly, teachers should be aware that only appropriate and relevant humor may yield positive outcomes; aggressive, disparaging humor especially when not related to the course materials most often, if not always, does much more harm than any good (see Wanzer et al., 2006). As Banas et al. (2011, p. 136) recommend "teachers should utilize humor that laughs with students rather than at them [all emphasis original]". Secondly, an optimal dosage of humor should be employed at the right points of instruction. Despite some recommendations in the extant research, I think teacher's intuition (and action research), well-seasoned with his/her declarative and practical knowledge, can be the best guide. Thirdly, sociocultural factors (e.g. age, gender, religion) may remarkably influence the way humor is produced, perceived, and projected (see also Zhang, 2005). Therefore, teachers should consider the sociocultural as well as educational norms before implementing humor in their classrooms. Otherwise, the affective and cognitive effects of humor might not be obtained. Fourthly, personality-based differences in humor should be taken into account (Martin et al., 2003). In employing different types of humor, teachers should not be forced. Comic video clips and cartoons can be a good starting point to use humor particularly for low-humor oriented teachers or in the settings where instructional humor is not fully welcomed yet (Banas et al., 2011). In addition, in-service teacher training courses and workshops with a focus on pedagogical humor as well as using new technologies might be of help in empowering teachers and learners to make the best use of classroom humor (Banas et al., 2011;Petraki & Pham Nguyen, 2016). Likewise, microteaching using (video-recordings of) successful instances of classroom humor can also aid teachers in implementing humor properly.
Within the purview of the study's limitations, it is recommended that interested researchers examine the implementation of teacher humor with a special focus on the teachers' individual differences as well as high-order educational norms and conventions. Also, an investigation into the interplay between teacher and learner humor, tracing the development of teacher humor through longitudinal studies, and a critical look at humor culture in language settings are fruitful areas of research in L2 (teacher) humor scholarship. Notes 1. Cook (1997) further distinguishes two types of fun or ludic language play: (a) play with language "form" (e.g. pun, neologism, alliteration, parallelism) and (b) play with language "meaning" or "semantic" play (e.g. teasing, parody, double-voicing). 2. FTA here is an umbrella term that I preferred to use for what the participants referred to as the moments when they had to confront a problematic situation in class (e.g. a learner's misbehavior) and they wanted to be, in a participant's words, "both sharp and soft" (i.e. criticizing the thing/person while maintaining the friendly atmosphere of the class). 3. The quoted words and phrases henceforth, unless otherwise indicated or implied, come from the respective teacher participants especially from their video-cued, follow-up interviews, reflecting on their humor use. 4. This faux-pas is probably rooted in the learner's erroneous blending of two Persian idioms, "to fly the elephant" and "to break the monster's horn" (both meaning roughly as doing an almost impossible job and/or confronting a difficult or hazardous situation).