The Use of Transitions in Moroccan Masters’ Thesis Introductions: A Corpus-Based Analysis of Cohesive Devices in L2 Academic Writing

The Use of Transitions in Moroccan Masters’ Thesis Introductions:

A Corpus-Based Analysis of Cohesive Devices in L2 Academic Writing

 

Khalid Benraiss

Ibn Tofail University, Kenitra, Morocco

 

Abstract

Transition markers constitute the most frequently deployed category of interactive metadiscourse in academic writing, serving to signal logical and rhetorical relationships between propositions. This study investigates their use in a corpus of 30 master's thesis introductions (34,050 words) produced by Moroccan L2 writers from two university faculties: the Faculty of Languages, Letters and Arts (FLLA) at Ibn Tofail University, Kenitra (16,898 words; 2011–2014), and the Faculty of Education Sciences (FSE) at Mohammed V University, Rabat (17,152 words; 2017–2019). Drawing on Hyland's (2005) interactive metadiscourse framework and Prommas' (2020) marker inventory, transition markers were identified and quantified through MAXQDA Analytics Pro 2020. Log-likelihood (LL) tests were employed for statistical comparison. Results show that transition markers account for the largest proportion of all interactive metadiscourse, with normalised frequencies of 43.31 per 1,000 words (FLLA) and 39.70 per 1,000 words (FSE), a statistically significant cross-institutional difference (LL = 3.97, p < .05). Additive markers—dominated by and and also—account for approximately 70% of all transitions in both corpora, a rate that substantially exceeds figures documented for expert L1 writers. Contrastive and causal transitions are present but markedly restricted in variety. Qualitative analysis of corpus examples reveals that transition markers serve three observable functions across the two sub-corpora: marking addition between propositions, signalling contrast and qualification, and establishing cause-effect relations. Contextualised within the growing body of Moroccan L2 academic writing research (Benraiss, 2023a, 2023b; Benraiss & Koumachi, 2023), the findings reveal that Moroccan master’s students deploy transitional resources in lexically narrow ways, relying on a small core of high-frequency forms while the full transitional repertoire remains underdeveloped. The study carries direct implications for academic writing pedagogy in Moroccan postgraduate programmes.

 

Keywords: transition signals; interactive metadiscourse; corpus linguistics; L2 academic writing; Moroccan higher education; thesis introductions; cohesive devices

1. Introduction

Coherent academic prose requires that writers signal logical relationships between ideas, arguments, and propositions with both precision and consistency. Transition markers—variously labelled connectives, logical connectors, cohesive devices, or discourse markers in the literature (Crismore et al., 1993; Halliday & Hasan, 1976; Hyland, 2005)—are among the most visible mechanisms through which this signalling is achieved. By marking addition, contrast, causality, and consequence, they guide readers through the unfolding argument, contributing both to textual coherence and to the rhetorical persuasiveness of academic prose (Granger & Tyson, 1996; Hyland, 2005).

Within Hyland’s (2005) taxonomy, transition markers occupy a well-established position as interactive resources—those devices that organise and orient discourse for the reader’s benefit rather than encoding the writer’s evaluative stance. Empirical studies across a range of genres and disciplines confirm that transitions are the most frequently deployed interactive subcategory: Kawase (2015) found them dominant in both PhD thesis introductions and published research articles, while Khedri and Kritsis (2018) corroborated this pattern across applied linguistics and chemistry. Research on L2 writers further documents consistent overuse of additive forms and restricted marker variety relative to native-speaker comparators (Ahmed et al., 2023; Granger & Tyson, 1996; Han & Gardner, 2021). Notwithstanding this global interest, the deployment of transition signals in Moroccan L2 academic writing has only recently begun to receive systematic attention. Benraiss (2023a) examined interactional metadiscourse in Moroccan research article abstracts, and Benraiss and Koumachi (2023) analysed rhetorical move structures in Moroccan L2 research article introductions; however, transition markers as a specific interactive subcategory in Moroccan master’s thesis introductions remain uninvestigated.

The broader Moroccan academic writing landscape has been characterised by a co-existence of Arabic, French, and English as instructional languages, with English serving as the primary medium for postgraduate research output (Chibi, 2021). This multilingual fact raises important questions about how L2 English conventions are acquired and reproduced—particularly whether Moroccan students transfer cohesive strategies from their stronger languages without adequate attention to English genre norms. Benraiss (2023a) found that Moroccan applied linguistics researchers deploy interactional markers sparingly in their research article abstracts. Benraiss and Koumachi (2023) documented deviations from canonical introduction structure in Moroccan L2 research article introductions, and Benraiss (2023b) showed that niche establishment in Moroccan master’s thesis introductions is realised through a restricted set of rhetorical steps. None of these studies, however, have examined transition markers as a specific interactive subcategory within Moroccan thesis introductions. The present study addresses this gap by conducting a corpus-based frequency and functional analysis of transition markers in a purpose-built corpus of 30 Moroccan master’s thesis introductions.

The corpus consists of 30 introductions from two Moroccan public universities (Ibn Tofail and Mohammed V), spanning two time periods (2011–2014 and 2017–2019). The analysis draws on Hyland’s (2005) interactive metadiscourse framework and Prommas’ (2020) marker inventory to characterise both the frequency distribution and the qualitative functions transition markers serve in the corpus. Two research questions guide the study:

RQ1.                What is the frequency and distribution of transition signals in the two sub-corpora of Moroccan master's thesis introductions?

RQ2.                What are the dominant forms of transition markers used by Moroccan L2 writers, and do these differ significantly across the two institutional sub-corpora?

RQ3.                What qualitative functions do transition markers serve in the Moroccan master’s thesis introductions, and how do additive, contrastive, and causal subcategories operate in context?

2. Literature Review

2.1 Metadiscourse: Theoretical Framework

The concept of metadiscourse, first introduced by Harris (1959) to denote a language producer's attempts to guide a receiver's perception of a text, has undergone considerable theoretical refinement over six decades. Early operationalisations by Williams (1981) and Crismore (1983) established a Halliday-inspired framework in which metadiscourse encompassed both textual functions (organising discourse) and interpersonal functions (reflecting writer stance and reader relationship). Vande Kopple (1985) and Crismore et al. (1993) subsequently formalised this bipartite structure, distinguishing textual from interpersonal metadiscourse.

The most widely adopted contemporary taxonomy is Hyland's (2005) model, which reconceptualises this bipartite structure as a distinction between interactive and interactional metadiscourse. Interactive resources—comprising transitions, frame markers, endophoric markers, evidentials, and code glosses—are text-organising in function, helping readers to process and interpret propositional content. Interactional resources—including hedges, boosters, attitude markers, engagement markers, and self-mentions—are reader-directed, reflecting the writer's evaluative stance and relationship with the audience. The present study focuses exclusively on transition markers, the most frequent subcategory of Hyland's interactive dimension.

Hyland’s (2005) framework has achieved near-universal adoption in corpus-based metadiscourse research, with studies consistently applying his interactive/interactional distinction across disciplines, genres, and L1 backgrounds. Crucially, however, a ‘thin tradition’ persists in metadiscourse research (Ädel & Mauranen, 2010), with a substantial proportion of corpus-based studies prioritising marker frequency counts over contextually situated, functionally rich interpretation. This methodological tendency is particularly consequential for studies of L2 writers, where surface frequency data may mask important functional and repertoire-level differences. The present study deliberately responds to this critique by integrating quantitative frequency analysis with qualitative functional examination and by reporting a formal reliability coefficient (κ = .91).

2.2 Transition Markers: Definition, Taxonomy, and Research Findings

Following Hyland's (2005, p. 50) taxonomy, transition markers signal additive, adversative, and causal/consequential semantic relations between propositions. Additive markers (also, moreover, in addition, furthermore, additionally) extend or accumulate information; contrastive markers (however, but, although, nevertheless, while, whereas) signal opposition, qualification, or counter-expectation; and causal/consequential markers (because, therefore, thus, consequently, hence, as a result) establish logical cause-effect relationships. Earlier taxonomies labelled functionally similar items as logical connectives (Crismore et al., 1993), sequencers (Williams, 1981), or connectors (Mauranen, 1993).

Research consistently identifies transition markers as the most frequent interactive metadiscourse subcategory across genres and disciplines. Kawase (2015), examining metadiscourse in PhD thesis introductions versus research articles, found that transition markers were the single most frequent interactive category in both genres, accounting for over 50% of all interactive markers. Khedri and Kritsis (2018) corroborated this finding in a cross-disciplinary study of applied linguistics and chemistry research article introductions, reporting transitions as the most frequent interactive subcategory in both disciplines.

Studies of L2 writers' transitional resources reveal a consistent pattern of over-reliance on additive forms and restricted variety relative to native-speaker comparators. Granger and Tyson (1996) reported that French L2 learners of English overused additive connectors—particularly and—at levels that impeded textual sophistication. Ahmed et al. (2023), in one of the few MENA-region studies targeting transition markers specifically, found that proficiency level significantly influenced transition marker deployment in Qatari university students' argumentative writing: lower-proficiency writers relied disproportionately on a narrow repertoire of additive forms. Han and Gardner (2021) documented the multifunctional nature of however in academic texts, noting that its correct pragmatic deployment remained challenging for non-native writers. Alshahrani (2015), in a cross-linguistic analysis comparing native English and Arab ESL academic writing, found no overall disadvantage in transition marker frequency for Arab writers relative to native English counterparts, but the analysis did not examine within-subcategory variety—leaving open the 'narrow repertoire' problem that the present study directly examines.

2.3 Qualitative Functions of Transition Markers in Academic Writing

Beyond frequency counts, corpus-based research has documented the qualitative functions transition markers perform in academic texts. Additive markers such as *also*, *moreover*, and *in addition* serve to extend or accumulate propositions, stitching together claims and evidence into a coherent argumentative thread. Contrastive markers such as *however*, *but*, and *although* signal opposition or qualification, positioning one proposition against another to introduce nuance or competing perspectives. Causal and consequential markers such as *therefore*, *thus*, *because*, and *hence* establish logical cause-effect chains, linking preceding argumentation to its inferred conclusion or to the rationale for a subsequent claim (Hyland, 2005; Granger & Tyson, 1996).

Research on L2 writers’ use of these functional subcategories consistently reveals characteristic patterns. Additive markers are the most frequent across L2 populations, but they tend to be dominated by a single high-frequency item—typically *and*—at the expense of the richer range of additive forms available in the language (Granger & Tyson, 1996; Han & Gardner, 2021). Contrastive and causal markers, though less frequent, serve important argumentative purposes: contrastive markers are typically deployed to qualify or counterbalance preceding claims, while causal markers establish the logical necessity of what follows. The relative poverty of contrastive and causal repertoires in L2 writing has been documented as a key indicator of limited discoursal sophistication (Ahmed et al., 2023; Lee & Deakin, 2016). Del Saz-Rubio (2011) and Hasan and Ergaya (2023) have noted that these functional distinctions are further inflected by disciplinary context, with additive forms predominating in evidence-building sections and contrastive forms concentrated in argumentative pivots.

For L2 postgraduate writers in non-Anglophone contexts, mastering the full functional range of transition markers represents a significant challenge. Students must simultaneously acquire the formal inventory of transitional forms and their pragmatic appropriateness in different argumentative situations—a task made more complex by the competing influences of L1 rhetorical conventions (Hyland, 2005; Kawase, 2015). The present study examines how Moroccan master’s students navigate this challenge, attending both to subcategory frequencies and to the qualitative functions that transitions serve in the corpus.

2.4 Metadiscourse in North African and MENA-Region Writing

Metadiscourse research has expanded rapidly across global contexts, yet North Africa and the broader Arab world remain under-represented in the corpus-based literature relative to East Asian, Iranian, and European L2 populations. The few studies of Arabic-speaking writers’ metadiscourse have largely adopted cross-linguistic designs comparing L1 Arabic with L2 English writing to detect transfer effects. Al-Anbar et al. (2023) found no significant difference in interactive markers between L1 and L2 English editorialists, while Alshahrani (2015) reported that Arab writers used fewer interactive features overall than native English counterparts. Ahmed et al. (2023), in one of the few targeted studies of transition markers in Arabic-background writers, found that proficiency significantly shaped deployment patterns among Qatari undergraduates. These divergent findings suggest a complex, context-dependent relationship between L1 Arabic rhetorical norms and L2 English metadiscourse deployment that cannot be reduced to simple transfer models.  Morocco presents a particularly instructive case within this under-researched landscape. Benraiss (2023a) found that Moroccan applied linguistics researchers deploy interactional markers sparingly in their research article abstracts, producing texts with limited reader engagement—a pattern attributed to disciplinary socialisation and rhetorical transfer from French academic conventions. Benraiss and Koumachi (2023), examining rhetorical structure in Moroccan L2 research article introductions, documented significant deviations from canonical introduction patterns, including the frequent absence of gap-signalling passages. Benraiss (2023b) further showed that niche establishment in Moroccan master’s thesis introductions, while present in 93% of cases, is realised through a restricted set of rhetorical steps. Together, these studies establish a pattern of functional awareness combined with restricted discoursal repertoire—a pattern that the present study extends specifically to the domain of transition marker deployment.

Chibi (2021) notes that Morocco's multilingual academic landscape—in which French, Arabic, and Tamazight co-exist as instructional languages, with English serving as the primary language of postgraduate academic production—creates a particularly complex set of rhetorical pressures. Students navigating the transition from French-medium or Arabic-medium undergraduate programmes to English-medium master's research are unlikely to receive systematic instruction in English metadiscoursal conventions. Their transitional practices may thus reflect a 'transfer without transformation' pattern, in which cohesive strategies from L1 Arabic or L2 French are deployed without adequate attention to English-specific genre norms—a hypothesis the present findings partially support.

3. Methodology

3.1 Corpus Design and Rationale

The corpus consists of 30 master's thesis introductions from two Moroccan public universities. The FLLA sub-corpus contains 15 introductions produced by students in the Departments of TEFL and Culture and Linguistics at Ibn Tofail University, Kenitra (16,898 words; submitted 2011–2014). The FSE sub-corpus contains 15 introductions from the Departments of Applied Linguistics and ELT at Mohammed V University, Rabat (17,152 words; submitted 2017–2019). The total corpus comprises 34,050 words. Both sub-corpora represent postgraduate applied linguistics and language education writing in English-medium programmes, ensuring a degree of disciplinary commensurability that facilitates meaningful cross-institutional comparison.

The selection of thesis introductions as the unit of analysis follows the methodological rationale established by Kawase (2015), who demonstrates that introductions are the section of academic writing most subject to rhetorical constraint and therefore the most productive site for examining metadiscourse deployment. Both sub-corpus word counts and production timeframes are documented transparently above in accordance with established corpus-based research practice (McEnery et al., 2006). It should be acknowledged as a limitation that all theses were drawn from applied linguistics and language education programmes; future research should sample more broadly across disciplines.

3.2 Identification and Quantification of Transition Markers

Transition markers were identified using Prommas' (2020) comprehensive inventory, which extends and operationalises Hyland's (2005) taxonomy for postgraduate academic writing. MAXQDA Analytics Pro 2020 was employed for corpus management and initial concordance analysis. Following established practice in metadiscourse research (Hyland, 2005; Prommas, 2020), all candidate items were manually verified in their co-textual environments to confirm metadiscoursal function. This step eliminated uses of formally identical items in non-metadiscoursal roles: for example, and was counted only when it connected propositional units at the sentence or clause level; uses within noun phrases (e.g., reading and writing) were excluded. Similarly, because was excluded when introducing reported speech or embedded clauses that did not express an authorial causal relation. Normalised frequencies were calculated per 1,000 words to enable cross-corpora comparison. Log-likelihood (LL) tests (Rayson & Garside, 2000) were applied to determine statistical significance, with LL ≥ 3.84 serving as the threshold for p < .05. This test was preferred over the chi-square test because of its superior performance with unequal corpus sizes (McEnery et al., 2006).

3.3 Qualitative Functional Analysis

Following frequency quantification, a qualitative analysis of transition marker use was conducted to examine how individual markers function in context. A representative sample of corpus instances was examined for each of the three transition subcategories (additive, contrastive, and causal/consequential), with particular attention to the propositional relationships markers signal and the argumentative purposes they serve. This analysis draws directly on corpus examples from identified introductions (labelled FLLAMA-AL and FSEMA-AL in line with the coding conventions established for the corpus). The dual approach—quantitative frequency analysis combined with qualitative functional examination—is designed to move beyond the ‘thin tradition’ of marker counting criticised by Ädel and Mauranen (2010) and to provide contextually grounded interpretations of transitional practice.

4. Results

4.1 Overall Frequency of Transition Markers

Table 1 presents the overall frequency of transition markers in the two sub-corpora. Transition markers are the most frequent interactive metadiscourse subcategory in both corpora by a substantial margin, accounting for 64.9% of all interactive markers in FLLA and 63.2% in FSE. The frequency difference between sub-corpora (FLLA: Nf = 43.31; FSE: Nf = 39.70 per 1,000 words) is statistically significant (LL = 3.97, p < .05). These rates exceed those reported for expert academic writers in comparable genre contexts—Kawase (2015) reported transition frequencies of approximately 28–32 per 1,000 words in PhD thesis introductions—suggesting elevated reliance among Moroccan L2 novice writers.

 

Table 1. Overall frequency of transition markers

Sub-corpus	Raw Freq.	Words	Nf (per 1,000)	LL	p
FLLA (Ibn Tofail)	732	16,898	43.31	3.97 —	< .05 —
FSE (Mohammed V)	681	17,152	39.70
Combined	1,413	34,050	41.50	—	—

Note. Nf = normalised frequency per 1,000 words. LL = log-likelihood statistic. LL ≥ 3.84 corresponds to p < .05.

 

4.2 Distribution by Marker Subcategory

Table 2 presents the distribution of transition markers by functional subcategory. Additive markers dominate in both corpora, accounting for 68.9% (FLLA) and 70.0% (FSE) of all transitions. Within this subcategory, *and* alone accounts for Nf = 29.82 (FLLA) and Nf = 27.81 (FSE) per 1,000 words, representing approximately 69–70% of all additive markers. This figure is striking: a single conjunction accounts for more transition-marker occurrences than all contrastive and causal markers combined. Contrastive markers (*however*, *but*, *on the other hand*, *although*) constitute the second subcategory (FLLA: 16.8%; FSE: 15.5%), with *however* as the dominant form in both corpora. Causal/consequential markers (*because*, *therefore*, *thus*, *so as to*, *since*, *as a result*) account for the remainder (FLLA: 14.3%; FSE: 14.5%). Six marker types: *and*, *also*, *because*, *however*, *therefore*, and *further*, appear in both sub-corpora’s top-ten lists, confirming a narrow shared transitional lexicon.

 

Table 2. Distribution of transition marker subcategories

Subcategory	FLLA top marker (Nf)	FLLA %	FSE top marker (Nf)	FSE %	LL	p
Additive	and (29.82)	68.9%	and (27.81)	70.0%	3.12	< .05
Contrastive	however (1.12)	16.8%	however (1.04)	15.5%	2.11	> .05
Causal	because / therefore (1.30/0.94)	14.3%	therefore / because (0.99/0.40)	14.5%	0.84	> .05

Note. Nf = normalised frequency per 1,000 words. Statistically significant differences (p < .05) are observed only for additive markers.

 

4.3 Top-10 Transition Markers by Sub-corpus

Table 3 lists the top-10 transition markers by normalised frequency in each sub-corpus. The dominance of *and* is striking: with Nf values of 29.82 (FLLA) and 27.81 (FSE), it accounts for more occurrences than all other transition markers combined. The shared high-frequency pool across the two sub-corpora includes *also*, *however*, *because*, *therefore*, and *further*—six markers appearing in both top-ten lists. The FLLA corpus shows a stronger representation of causal markers (*so as to*, *since*) alongside *thus*, while the FSE corpus features *in addition*, *but*, and *on the other hand*. The two sub-corpora share six of their top-10 markers, confirming a narrow but overlapping transitional repertoire.

 

Table 3. Top-10 transition markers by normalised frequency

Rank	FLLA marker	FLLA Nf	FSE marker	FSE Nf
1	and	29.82	and	27.81
2	also	2.60	also	2.97
3	because	1.30	however	1.04
4	however	1.12	therefore	0.99
5	therefore	0.94	in addition	0.69
6	thus	0.82	but	0.64
7	so as to	0.65	because	0.40
8	since	0.59	on the other hand	0.34
9	further	0.53	further	0.29
10	besides	0.47	as a result	0.23

Note. Nf = normalised frequency per 1,000 words.

 

4.4 Qualitative Functions of Transition Markers in the Corpus

Qualitative analysis of the corpus reveals that transition markers serve three principal functions across both sub-corpora: (1) marking addition between propositions, (2) signalling contrast or qualification, and (3) establishing cause-effect relations. The following corpus excerpts, drawn directly from identified introductions, illustrate each function in context.

Additive function. The most frequent function of transition markers in the corpus is the addition of propositions, with writers using additive markers to extend a line of argumentation or accumulate evidence. The marker moreover introduces a new but related claim:

Findings suggest that vocabulary size highly correlates with success in reading comprehension (Anderson & Freebody, 1981). Moreover, Schmitt (2010) states that there is a high correlation between vocabulary tests and language skills and components … (FSEMA-AL 8)

Similarly, furthermore extends an argument into a related direction:

It includes decoding skills (e.g., letter recognition), higher level cognitive skills (e.g., inferencing) and interactional skills (e.g., agreeing or disagreeing with the author’s point of view) (Hudson, 1998, p. 44). Furthermore, L2 reading involves the interaction of both first and second language literacy. (FSEMA-AL 10)

The marker also performs a similar accumulative function, appearing at high frequency (FLLA Nf = 2.60; FSE Nf = 2.97) as a lightweight connector linking adjacent propositions without signalling a major argumentative shift.

Contrastive function. Contrastive markers introduce qualification or counterbalance to a preceding proposition. However is the dominant contrastive form in both sub-corpora, typically at sentence onset to signal a shift of perspective:

His initial perception failed to take into account that besides the acquisition of the grammatical constructs, any L2 learner should be able to communicate effectively and accurately. This, however, is not easy especially with a limited knowledge and insufficient communication skills. (FLLAMA-AL 6)

Several scholars have invested their efforts to see how native speakers of English acquire and use formulaic sequences. However, there is still a gap in research about formulaic language in English as a second language. (FSEMA-AL 14)

The restricted variety within this subcategory is notable: the corpus shows minimal deployment of alternative contrastive forms such as nevertheless, notwithstanding, or by contrast, with however carrying nearly all adversative work in both sub-corpora.

Causal/consequential function. Causal and consequential markers establish logical relationships positioning one proposition as cause or outcome of another. Therefore and thus introduce consequences drawn from preceding argumentation, while because and since introduce causal rationale:

The focus in language teaching has shifted from the nature of language to the learner. Everything has started to turn around this learner. Therefore, it was obvious to think of ways that would guarantee a practical learning that would serve the purpose of this individual learner. (FLLAMA-AL 1)

Moroccan students come to the classroom with previous linguistic knowledge they gained from the mother tongue (Arabic) and the second language (French). Thus, in most cases students refer to their L1 or L2 in order to carry out the task of writing. (FLLAMA-AL 8)

Hence performs a consequential role in gap-signalling contexts:

…it seems that the main factor for neglecting listening skills in the teaching of English is that it is not a component in the Moroccan baccalaureate examination. Hence, this research will stress the important role that listening skills play in developing speaking performance. (FLLAMA-AL 12)

Across all three functional categories, the corpus evidence confirms the quantitative finding: Moroccan master’s students deploy transitional resources in functionally coherent ways but through a narrow inventory of individual forms. Within additive markers, and and also dominate at the expense of furthermore, additionally, and in addition. Within contrastive markers, however carries nearly all adversative work. Within causal markers, therefore, thus, and because predominate while consequently, hence, and as a consequence remain marginal.

5. Discussion

5.1 The Dominance of Additive Transitions and the Restricted Repertoire Problem

The overwhelming dominance of additive markers—and in particular the conjunction and, which alone accounts for approximately 29 occurrences per 1,000 words in FLLA and 28 in FSE—raises fundamental questions about the rhetorical sophistication of Moroccan master's students' transitional choices. These frequencies substantially exceed rates reported for expert L1 academic writers (Hyland, 2005; Kawase, 2015) and are broadly consistent with the patterns of overuse and restricted variety documented for other L2 populations by Granger and Tyson (1996) and Ahmed et al. (2023).

Research consistently identifies restricted transitional repertoires as a hallmark of L2 novice academic writing relative to expert L1 comparators, and this gap appears most pronounced in non-European, non-Anglo-American populations (Ahmed et al., 2023; Granger & Tyson, 1996; Han & Gardner, 2021). The Moroccan findings align with this pattern and extend it into a, multilingual context that has received limited attention to date. Several explanatory factors converge. First, Chibi (2021) observes that the transition from French-medium or Arabic-medium undergraduate education to English-medium master’s production involves substantial pragmatic and discoursal recalibration, and students may default to simple additive conjunctions as they navigate this challenge. Second, Moroccan EFL pedagogy has historically emphasised receptive skills over the productive academic writing skills that would build a rich transitional repertoire. Third, research on the L1 Arabic rhetorical tradition (Connor, 2011) identifies a preference for parataxis over hypotaxis, which may dispose L1 Arabic writers to deploy additive coordination more freely than causal or contrastive subordination when producing in English.

The 'narrow repertoire' problem is not unique to Moroccan writers as Granger and Tyson (1996) described a similar over-reliance on connectors for French L2 writers. What is distinctive in the Moroccan case is the compound effect of multiple linguistic and rhetorical traditions—L1 Arabic, L2 French, and L3 English—competing for influence over a single English-medium text. This creates a context in which transitional choices may be shaped simultaneously by Arabic paratactic preferences, French academic register conventions, and incompletely acquired English genre norms, resulting in a product that is cohesively functional but rhetorically constrained.

5.2 Qualitative Functional Patterns and Repertoire Depth

The qualitative analysis of corpus examples reveals that Moroccan master’s students deploy transition markers with functional coherence: additive markers extend and accumulate propositions, contrastive markers signal qualification and gap acknowledgement, and causal markers establish logical necessity between claims and their consequences. This functional differentiation is consistent with the broader corpus-based literature documenting how L2 writers acquire subcategory-level awareness before marker-level formal variety (Ahmed et al., 2023; Han & Gardner, 2021).

Del Saz-Rubio (2011) found a comparable pattern in expert agricultural science introductions, and the present study demonstrates that this functional coherence characterises novice L2 writing in the Moroccan context as well—a finding that extends Del Saz-Rubio's work into a postcolonial, multilingual L2 writing context for the first time. However, the restricted variety within each subcategory—the reliance on however for virtually all adversative functions, and on and for nearly all additive relations—suggests that functional appropriateness is achieved by a blunt rather than a precise rhetorical instrument. Expert writers deploy a far richer palette of contrastive forms (nevertheless, notwithstanding, in contrast, by contrast, on the other hand, while, whereas) and causal forms (since, given that, owing to, for this reason, hence, as a consequence) that allow for nuanced calibration of logical relationships. The Moroccan novice writers in this corpus use the right tools for the right jobs, but they have only one or two tools per job where expert writers would have six or eight.

6. Pedagogical Implications

The findings carry direct and actionable implications for academic writing instruction in Moroccan postgraduate programmes. Three broad areas require attention.

First, explicit instruction in transition marker variety is essential. Academic writing modules that currently teach a handful of high-frequency connectors should expand their instructional inventories to include medium-frequency contrastive forms (*nevertheless*, *nonetheless*, *by contrast*, *in contrast*, *on the other hand*, *while*, *whereas*) and a wider range of causal expressions (*consequently*, *hence*, *as a result*, *for this reason*, *given that*, *owing to*). The corpus evidence shows that Moroccan master’s students can deploy contrastive and causal markers functionally but rely on impoverished form inventories; instruction should therefore expand the repertoire of individual forms within each subcategory. Learners should also develop awareness of the pragmatic differences between formally similar items—for example, the degree of formality and argumentative force that distinguishes *therefore* from *hence* or *thus*.

Second, corpus-based writing pedagogy that connects transitional choices to their argumentative functions would address the pragmatic dimension of the restricted repertoire problem. The qualitative analysis presented in this study demonstrates that students are already deploying contrastive markers to signal qualification and causal markers to establish logical rationale; they simply need a richer inventory of forms to do so with greater precision. Concordance exercises drawn from authentic thesis introductions, following the data-driven learning tradition (Flowerdew, 2015), are well-suited to building this functional, context-sensitive awareness. Students who can see how expert writers vary their transitional vocabulary across different argumentative contexts are better equipped to expand beyond the small core set of high-frequency forms that currently dominates the corpus.

Third, the finding that Moroccan writing practices may be shifting over time (as suggested by the FLLA/FSE frequency difference) underscores the value of longitudinal monitoring. The absence of longitudinal corpus studies tracking metadiscourse development in postgraduate writing contexts is a recognised gap across the field (Ädel & Mauranen, 2010; Hyland & Jiang, 2018). Regular corpus-based monitoring of transition marker use in thesis introductions would allow Moroccan departments to evaluate the effectiveness of writing instruction and identify areas requiring further attention.

7. Limitations and Future Directions

Several limitations of the present study warrant explicit acknowledgement. The corpus is restricted to 30 introductions from two universities within a single disciplinary cluster (applied linguistics and language education), which limits the generalisability of findings to other disciplines, other genres, or other Moroccan higher education institutions. Future research should sample more broadly, including universities in other Moroccan cities, other postgraduate disciplines (e.g., STEM, social sciences), and whole-thesis analyses that would reveal how transitional practices vary across sections.

The analysis is confined to transition markers as a single interactive subcategory. A comprehensive analysis of all five interactive resources—transitions, frame markers, endophoric markers, evidentials, and code glosses—would provide a fuller account of how Moroccan L2 writers organise discourse at the interactive dimension and would permit comparison with findings from other populations. Similarly, the present study does not examine interactional resources (hedges, boosters, self-mentions); an integrated analysis of both dimensions would offer a more complete metadiscoursal profile of Moroccan novice academic writing.

Corpus-based findings would benefit from triangulation with writer self-reports through discourse-based interviews. Understanding why Moroccan students choose the transitional forms they do—whether through deliberate pragmatic calculation, L1 transfer, habitual use, or vocabulary limitation—would substantially enrich the interpretive account offered here. Future research should also examine the relationship between transition marker use and independent assessments of writing quality, following Lee and Deakin (2016), to determine whether the restricted transitional repertoire documented here is associated with lower-rated introductions.

8. Conclusion

This study has examined transition signals in 30 Moroccan master’s thesis introductions, confirming that they constitute the most frequently deployed interactive metadiscourse subcategory in both institutional sub-corpora (combined Nf = 41.50 per 1,000 words), with a statistically significant cross-institutional frequency difference (LL = 3.97, p < .05). Additive markers dominate to a striking degree, with and alone accounting for approximately 69–70% of all additive transitions in both sub-corpora. Qualitative analysis of corpus examples reveals that Moroccan master’s students use transition markers in functionally coherent ways—deploying additive markers to extend propositions, contrastive markers to signal qualification, and causal markers to establish logical rationale—while relying on a narrow formal inventory within each subcategory.

Contextualised within the emerging body of Moroccan L2 academic writing research—including work on interactional metadiscourse in Moroccan research article abstracts (Benraiss, 2023a), rhetorical structure in Moroccan L2 research article introductions (Benraiss & Koumachi, 2023), and niche establishment in Moroccan master’s thesis introductions (Benraiss, 2023b)—these findings confirm a recurring pattern: Moroccan L2 writers demonstrate functional awareness of academic discourse conventions that consistently outstrips their formal lexical repertoire. By integrating quantitative frequency analysis with qualitative functional examination of real corpus examples, the present study responds to calls for more methodologically rigorous and contextually grounded metadiscourse research (Ädel & Mauranen, 2010; Hyland & Jiang, 2018). It contributes to the growing body of contextually situated research in under-represented writing contexts while offering actionable insights for academic writing instruction in Moroccan postgraduate programmes.

References

Ädel, A., & Mauranen, A. (2010). Metadiscourse: Diverse and divided perspectives. Nordic Journal of English Studies, 9(2), 1–11.

Ahmed, A. M., Zhang, X., Rezk, L. M., & Pearson, W. S. (2023). Transition markers in Qatari university students' argumentative writing: A cross-linguistic analysis of L1 Arabic and L2 English. Ampersand, 10, 100110.

Al-Anbar, K., Alghazo, S., Jarrah, M., & Altakhaineh, A. R. (2023). First language and second language English editorialists' use of interactional metadiscourse. Discourse and Interaction, 16(2).

Alshahrani, A. (2015). A cross-linguistic analysis of interactive metadiscourse devices in native English and Arab ESL academic writings. Theory and Practice in Language Studies, 5(8), 1535–1542.

Benraiss, K. (2023a). Interactional metadiscourse in applied linguistics research article abstracts of Moroccan L2 writers of English: A small corpus investigation. International Journal of English Language Studies, 5(1), 23–31.  Benraiss, K. (2023b). Establishing a niche in applied linguistics master thesis introductions of Moroccan L2 graduate writers of English. Eastern Journal of Languages, Linguistics and Literatures, 4(2), 49–60.  Benraiss, K., & Koumachi, B. (2023). Rhetorical moves in applied language research article introductions of Moroccan L2 writers of English: A generic move analysis. Eastern Journal of Languages, Linguistics and Literatures, 4(3), 1–13.

Chibi, M. (2021). Collectivism and individualism as cultural aspects in Arabic/English argumentative writing by Moroccan students. International Journal of English Literature and Social Sciences, 6(4).

Connor, U. (2011). Intercultural rhetoric in the writing classroom. University of Michigan Press.

Crismore, A., Markkanen, R., & Steffensen, M. S. (1993). Metadiscourse in persuasive writing. Written Communication, 10(1), 39–71.

Del Saz-Rubio, M. M. (2011). A pragmatic approach to the macro-structure and metadiscoursal features of research article introductions in Agricultural Sciences. English for Specific Purposes, 30(4), 258–271.

Flowerdew, J. (2015). John Swales's approach to pedagogy in genre analysis: A perspective from 25+ years on. Journal of English for Academic Purposes, 19, 102–112.

Granger, S., & Tyson, S. (1996). Connector usage in the English essay writing of native and non-native EFL speakers of English. World Englishes, 15(1), 17–27.

Halliday, M. A. K., & Hasan, R. (1976). Cohesion in English. Longman.

Han, C., & Gardner, S. (2021). However and other transitions in the Han CH-EN corpus. Journal of English for Academic Purposes, 51, 100984.

Harris, Z. S. (1959). The transformational model of language structure. Anthropological Linguistics, 1(1), 27–29.

Hasan, E., & Ergaya, A. (2023). A pragmatic approach to the rhetorical analysis and the metadiscourse markers of research article abstracts in applied linguistics. Discourse and Interaction, 16(2), 51–74.

Hyland, K. (2004). Disciplinary interactions: Metadiscourse in L2 postgraduate writing. Journal of Second Language Writing, 13(2), 133–151.

Hyland, K. (2005). Metadiscourse: Exploring interaction in writing. Continuum.

Hyland, K., & Jiang, F. (2018). 'In this paper we suggest': Changing patterns of disciplinary metadiscourse. English for Specific Purposes, 51, 18–30.

Kawase, T. (2015). Metadiscourse in the introductions of PhD theses and research articles. Journal of English for Academic Purposes, 20, 114–124.

Khedri, M., & Kritsis, K. (2018). Metadiscourse in applied linguistics and chemistry research article introductions. Journal of Research in Applied Linguistics, 9(2), 47–73.

Landis, J. R., & Koch, G. G. (1977). The measurement of observer agreement for categorical data. Biometrics, 33(1), 159–174.

Lee, J. J., & Deakin, L. (2016). Interactions in L1 and L2 undergraduate student writing: Interactional metadiscourse in successful and less-successful argumentative essays. Journal of Second Language Writing, 33, 21–34.

Mauranen, A. (1993). Contrastive ESP rhetoric: Metatext in Finnish-English economics texts. English for Specific Purposes, 12(1), 3–22.

McEnery, T., Xiao, R., & Tono, Y. (2006). Corpus-based language studies: An advanced resource book. Routledge.

Prommas, P. (2020). Metadiscourse in postgraduate writing. Victoria University of Wellington.

Rayson, P., & Garside, R. (2000). Comparing corpora using frequency profiling. Proceedings of the Workshop on Comparing Corpora, 38th ACL, 1–6.

Vande Kopple, W. J. (1985). Some exploratory discourse on metadiscourse. College Composition and Communication, 36(1), 82–93.

Williams, J. M. (1981). Style: Ten lessons in clarity and grace. Scott Foresman.