Use of linking words in argumentative written texts Comparison between L1 and L2

Introduction

This research stems from an observation, as foreign language teachers, when correcting texts written by university students with low or intermediary levels: some mistakes related to textual organisation, cohesion or choice of discourse features seem not to be related to L2 limitations only, but rather to crosslinguistic transfer of uses or routines already present in L1i. As an attempt to adapt our teaching practice to those difficulties in both languages, an ad-hoc L1 (Spanish) and L2 (English) corpus was compiled among first year university students who had not received explicit instruction on this topic, which would enable to determine our starting point, i.e. their previous knowledge.

Spanish Secondary Education and Baccalaureate curricula attach particular importance to argumentation, beyond language subjects, since it will be needed in multiple spheres in life. In spite of this, many studies highlight the fact that education programmes do not emphasize the promotion of argumentative strategies accordingly or, if they do, these are not successfully acquired during preuniversity stages (Melero & Gárate, 2013; Ferretti & Graham, 2019) while others show similar concerns regarding university students (Fuentes-Román & Farlora-Zapara, 2019; Rapanta & Macagno, 2019).

As it is well known, linguistic competence in our L1 will affect any other language we learn, and, similarly, metalinguistic awareness will play an important role to extrapolate certain abilities, which makes explicit teaching of linguistic concepts necessary at different stages in education, in line with learners’ cognitive maturity. A series of studies have highlighted the relevance of crosslinguistic transfer when we add knowledge learnt in an L1 to an L2 (Odlin, 1989), but also from the L2 to the L1 or between several foreign languages (Forbes & Fisher, 2020). Tankó’s (2004) research with Hungarian university students reveals that this group of advanced English learners uses a higher number of linking words than native speakers do, although with a poorer repertoire. The author points at the exposure of these learners to a great number of academic texts during their studies but also to the emphasis when teaching linking words among the main reasons for those findings. In this same line, the study by Carrió-Pastor (2013) analyses linking words in academic texts written in English, comparing native speakers’ productions to those of English language learners with Spanish as their L1. The author noted a more frequent use of linking words by native speakers, who resort to them as a discourse strategy to guide the reader. Cotos (2014) also shows evidence of overuse, misuse, or underuse of different types of discourse elements when comparing texts written by English native speakers and English language learners.

Transfer can be positive, and thus enable learning, or negative such as under- or overuse of certain features, calques or misunderstandings. Excessive importance is frequently attached to the learner’s linguistic and metalinguistic knowledge in their L1, assuming it will enable the comprehension of similar aspects in the L2 and trusting that crosslinguistic transfer will eventually happen somehow (Forbes, 2020), although, as rightly noted by Forbes & Fisher (2020), such transfer can be limited and strategies used in the L1 might not be effective in the L2. For this reason, a pedagogic approach that considers both the curriculum and its different languages from a crosscurricular perspective can be particularly helpful for learners. To mention an example, approaches such as translanguaging, more and more frequent in plurilingual educational settings, attempt to take advantage of learners’ full linguistic repertoire in order to promote content and language acquisition, with explicit tasks such as cognate comparison (Cenoz et al., 2021).

We will hereinafter present a theoretical framework on argumentation (what it consists of and how it is implemented in preuniversity curricula) and linking words. Due to its limitations and in order to facilitate contrastive analysis between the languages involved, this study focuses only on those linking words used at the beginning of the paragraph. In addition to a high frequency of linking words which serve to organize the discourse, argumentative texts are characterized by a high frequency of epistemic and stance markers used to express the writer’s opinion, and, for that reason, these elements in initial position have also been considered. We will next describe the aims and method, results, conclusions and possible implications of this study for language teaching.

Argumentation, linking words and preuniversity curricula

Argumentation aims at persuading the addressee about an opinion or point of view on a given topic, and at supporting this view different reasons are provided. This way, argumentative texts have as their main aim “to refute one or several opinions designated in an explicit or implicit manner” in which the concepts “are interrelated through explicit internal causality relationships, so that they are built based on linking words of a logic nature” (translated from Bustos-Gisbert, 1996, p. 108-109). Therefore, an analysis and reflection ability is needed both to understand the premises the text is based on and to argument logically, a skill that is improved from childhood, and especially during adolescence.

The use of spoken argumentation emerges in early childhood when reasons to justify decisions start to be produced (Ferretti & Graham, 2019) and between 10 and 12 years the ability to use consistent and cohesive devices in our narratives is completed (Serra et al., 2000, p. 520). Nevertheless, explicit instruction during adolescence is necessary to develop critical thinking abilities, especially those related to argumentation, for instance through tasks that add in metacognitive reflection to argumentative practice (Felton, 2004). Table 1 shows references to recognition and use of both argumentation and cohesive devices in Royal Decree 1105/2014, which establishes the curriculum for Compulsory Secondary Education (CSE) and Baccalaureate in Spain. Different features describing subjects in its Appendix I have been considered, namely contents, assessment criteria and assessable learning standards. In the case of linguistic subjects there is a distinction between comprehension and production.

As shown in table 1, in both educational stages (Secondary Education and Baccaulareate), and in all L1, L2 and non-linguistic subjects (which could be taught in the L2 in bilingual programmes) reference is made to argumentation. Its remarkable presence in a varied range of subjects highlights the relevance of this ability to express and justify opinions both orally and in written texts, in any field, and, as a consequence, emphasizes learners’ need to finish these stages having acquired specific discourse strategies. In linguistic subjects, comprehension and production consider different text types and their structure. At the discourse level, accuracy in cohesive devices at textual level is present as a content from the initial stage in Secondary Education in the subject Spanish Language and Literature. The pragmatic features of argumentative texts appear in the Baccalaureate stage.

Another reference document in the case of L2 in the Common European Framework of Reference for Languages (CEFR), with descriptors that place the ability to use linking words to relate ideas and justify opinions in between the intermediate levels- B1 or threshold and B2 or vantage. Thus, among the descriptors for B1 ( CFER, 2001, p. 24) we can find “Can produce simple connected text on topics which are familiar or of personal interest,” while in B2 ( CFER, 2001, p. 241), “Can express news and views effectively, and relate to those of others” or “Can use a variety of linking words to mark clearly the links between ideas”. It should be noted that the Companion Volume to the CFER (Council of Europe 2020) explicitly adds -in B1 level- the reading comprehension of the main idea and conclusions of argumentative texts, and -within the written production- the synthesis and evaluation of information or arguments.

As can be seen, one of the necessary elements for learners to reach discourse cohesion, key to link the different sentences that make up a text, relating ideas coherently, lies on the knowledge of “linking words”, “textual organisers” or “connectors,” which help to mark existing links between sentences and facilitate comprehension. Due to their importance, they often occupy prominent positions within the sentence and are commonly placed at the beginning of the paragraph.

Among the different labels used to refer to this type of markers we can find conector as a conventionalised term in Spanish (Domínguez-García, 2006; Fuentes-Rodríguez, 2009), often as a subcategory within discourse markers (Martín-Zorraquino & Portolés-Lázaro, 1999; Pons-Bordería, 2001). In English they are referred to as ‘conjuncts’ (Quirk et al., 1985), ‘discourse markers’ as a subcategory within ‘pragmatic markers’ (Fraser, 1996), ‘linking adverbs/adverbials’ (Biber et al., 1999), ‘connective adjuncts’ (Huddleston & Pullum, 2002) or ‘sentence connectors’ (Carrió-Pastor, 2013), to mention just a few. For the catalogues of linking words in Spanish different sources were used as reference, namely Martín-Zorraquino & Portolés (1999); Domínguez-García (2006) and Fuentes-Rodríguez (2009); while Quirk et al. (1985); Biber et al. (1999); Huddleston & Pullum (2002) and Carrió-Pastor (2013) were considered in the case of English.

Other ways of introducing the paragraph include the so-called stance or viewpoint markers. Biber et al. (1999, p. 855) identify as ‘epistemic stance adverbials’ elements such as in our view, from our perspective, in my opinion, whose function consists in marking the point of view or the perspective from which the proposition is true. Ohlrogge (2009), who uses the label “personal stance markers” when referring to expressions such as I think, shows this type of sequences among the most frequent formulaic expressions in written exams produced by a group of intermediate learners. In a study on argumentative texts written by university learners of English as a Foreign Language, Salazar & Verdaguer (2009) note that think is the most common lexical verb among Spanish participants in the corpus and also the most frequent one to express modality, an overuse that the authors link to beginners. The use of opinion verbs is recurrent in argumentative spoken discourse both to show security and certainty on what is said and to do the opposite: to contribute to soften the assertion and favour the nuance of subjectivity in the statement, as suggested by studies in English using I think (Aijmer, 1997) and in Spanish using suponer, pensar or creer (Brenes-Peña, 2015).

Given the huge differences regarding definitions and catalogues in the studies on linking words, this term will be used here in a broad sense. Following Fuentes-Rodríguez (2009, p. 15), we will consider “those which are fixed or in an advanced grammaticalization stage” and only those which are placed at the beginning of the paragraph, where they display an evident discourse function, including those with mistakes, both in Spanish (L1) and in English (L2). Additionally, this criterion allows the inclusion of periphery elements with an obvious pragmatic or discourse intention in spite of not adhering to standard style guidelines, such as pero (‘but’) at the beginning of the paragraph.

Objectives

The objectives of this study are (a) to observe and classify the use of linking words and stance markers between paragraphs in argumentative texts written by Spanish speakers in their L1 and in their L2 (English); (b) to detect frequent errors related either to the collocations of the different connectors or to discursive aspects; (c) to carry out a contrastive analysis of the linking words and stance markers in both languages to identify possible correspondences and indications of linguistic transfer; and (d) to check students’ competence in the use of these discursive strategies at the beginning of their university stage.

Method

Results

This section presents a catalogue of the linking words and stance markers located at the beginning of the paragraph in the texts written by the participants in L1 and L2. Their use is described, both quantitative and qualitatively, and some similarities found between both languages are exemplified.

Figures 3 and 4 show the presence or absence of connectors at the beginning of the paragraph in the texts written in both L1 and L2. In both cases they are most frequent in the last paragraph (P4), followed by the third (P3) and the second one (P2) with very similar figures, a number which becomes slightly higher in English in the second one and somewhat higher in Spanish in the last ones:

Figure 5 shows the types of linking words in the paragraph where they appear in the texts produced in English.

Similarly, figure 6 shows the type of markers in each paragraph in Spanish.

Table 3 details the catalogue of connectors in the texts in Spanish, classified by type of marker. The columns are divided into the subgroups that first wrote the text in Spanish (L1-L2) and those who wrote it first in English (L2-L1). As can be seen, the type of linking words used is similar, although the repertoire of the group that wrote first in their mother tongue (with a higher number of participants) is slightly greater:

Table 4 shows the catalogue of forms in the texts written in English. Compared to the texts in Spanish, both the lower presence of consecutive linking words and the absence of those of addition stand out in this case. Also striking is the large number of connectors used erroneously in English for the structure On the one hand/On the other hand:

As table 4 shows, some linking words that serve to organize discourse ideas in adjacency pairs (On the one hand... On the other hand...) are likely to occur in rather unacceptable combinations, something that can be found in paragraphs with other types of lexical or grammatical errors, as in (1). However, the errors in these correlational expressions can also be observed in texts with greater grammatical accuracy, as (2). Similarly, certain combinations found in Spanish like (3) might seem unnatural and give the impression of being a calque of the English version, even when the text in Spanish was written first:

An occurrence of but in initial position is found in example (4). The fact that but is used again in the same paragraph might reveal a certain lack of resources to indicate opposition, although a similar use occurs in Spanish, in (5):

As for the use of stance markers, as already pointed out, they are very frequent in argumentative texts. Students thus use them to express their point of view at the beginning of the paragraph, sometimes in combination with other markers or with linking words. In order to establish a comparison that is as symmetrical as possible, in the following figures and tables, those expressions that are not at the beginning of a paragraph or after a linking word or another stance marker opening the paragraph have not been taken into account.

As shown in tables 5 and 6, similar total numbers of stance markers are found in both languages: 66 in English and 60 in Spanish, although there is greater variety in the participants’ L1, with four different epistemic markers compared to one in the texts in English. The variety of forms in the stance markers does not always obey criteria of correctness or acceptability:

In this case, in order to simplify the table, spelling errors have not been taken into account and some forms such as for me and my opinion have been included despite not being idiomatic when they clearly serve to express the writer’s viewpoint.

The frequency of this type of structures is higher in English than in Spanish, possibly due to the limitations offered by their L2 skills. Studies like those by Salazar & Verdaguer (2009) or Basturkmen & Von Randow (2014) indicate that expressions of belief or personal evaluation such as I think in argumentative texts are more common in English learners at lower levels, which could also explain the difference from their native language. Despite their prominent position at the beginning of the paragraph, these elements contribute to a greater degree of subjectivity in the statement and to the writer’s insecurity being perceived. The presence of epistemics is frequently preceded by some kind of linking word, as in (6). In addition, a frequent -and redundant- combination of stance markers with epistemic markers is observed both in L1 and L2, as shown in examples (7)-(9):

It should be noted that the same person wrote examples (8) and (9), first in English and then in Spanish. Strikingly the structure in Spanish seems almost a literal translation from English, including even the first-person pronoun. These cases (see also example (3)) could exemplify the crosslinguistic transfer from L2 to L1 in formal written texts, for which explicit training is often received in L2 (Tankó, 2004).

Discussion and conclusions

Although the participating students were divided into two groups to avoid biases among those who wrote first in Spanish and then in English, the final catalogues of markers are almost identical in both groups, so this factor does not seem to have had much weight given that texts were short. In relation to the first objective stated, the correctness and variety of linking words is, obviously, greater in L1 (the greater lexical availability in Spanish is evident above all in the use of consecutives in the last paragraph). However, quite a few similarities can be seen in the selection of linking words used when writing in L1 and in L2:

As for the structures used at the beginning of the paragraph to indicate the writer’s opinion or stance, there is a greater variety of epistemic and stance markers in Spanish, with very similar total numbers in the latter. The epistemic ones are more frequent in English, particularly in the first paragraph, where students with lower levels may resort to I think due to the lack of knowledge of other more adequate structures to start the text. Thus, lacking other resources to introduce the argument, they limit themselves to reaffirming their position or point of view. Although the repertoire of linking words and epistemic markers in the L1 is richer, the frequency of exposure or the more explicit way of teaching them in the L2 might end up interfering with the production of written texts in their L1. Likewise, the fact of having more limited L2 skills leads to an overuse of simpler conventionalized structures (Tankó, 2004).

One of the aims of this work was to verify the students’ competence in the use of these discourse strategies at the beginning of their university stage. As shown through the summary of the contents related to the production of argumentative texts in Royal Decree 1105/2014 in Table 1, argumentation is strongly present in the Secondary Education and Baccalaureate curricula. However, this emphasis on the legislative framework can lead to misunderstanding the knowledge that students actually hold. Students are expected to know how to correctly use linking words and cohesive devices in argumentative texts and most of them are able to do so both in Spanish and in English, although with many errors in the L2, despite having studied it for many years, and with certain limitations in the L1.

Although the study is conditioned by the heterogeneity of levels, this diversity may not be an isolated fact in other university classrooms or in other academic levels, which leads to the need to rethink how to teach these structures in the L2, focusing on their character as fixed formulae (on the one hand… on the other hand) and establishing a comparison with Spanish.

Other objectives in this study were to detect frequent errors related either to the collocations of the different linking words or to discursive aspects, and to identify possible correspondences and indicators of linguistic transfer, which have been exemplified in the results section. Taking into account the relevance of crosslinguistic transfer (Forbes, 2020), it is important to focus on the explicit teaching of metacognitive strategies that allow such transfer to be carried out and to promote a multilingual methodological approach to develop written expression techniques. Possibly due to the need to reflect on linguistic aspects in the learning process of an L2, studies such as Forbes & Fisher’s (2020) point to a greater evidence of transfer from L2 to L1 than vice versa in writing strategies. It is important not to consider that prior knowledge will be homogeneous among students, so reviewing metalinguistic aspects in their L1 could be very useful. In this sense, activities such as those framed within translingualism (Cenoz et al., 2021), which usually takes advantage of the existence of cognates in first and foreign languages, can serve to highlight both elements that are common and those that are different in the languages known by the students, establishing a direct relationship between linguistic functions and the way of expressing them.

This work focused on elements such as linking words and stance markers that are highly frequent in the examined texts, which has allowed us to establish comparisons between the two languages, although other types of less fixed structures can perform the same cohesive function between paragraphs. In the future, with the aim of delving into this study, these structures could be taken into account, as well as the possible influence of certain variables of the participating students (such as their level of previous qualifications) on the results.

As a complement to the situation herein described, it would be important to carry out a didactic intervention involving the teachers of the two languages analysed so as to show effective writing strategies in both of them- containing contrastive aspects and enhancing the crosslinguistic transfer of discursive, textual cohesion and organization elements. Offering a wider range of contextualized connectors in both comprehension and production activities could contribute to an increase in students’ repertoire. Agreeing with Salazar & Verdaguer (2009) it is important to draw students’ attention to the variety of lexical verbs that can express modality, as well as to encourage greater use of impersonal structures to the detriment of more subjective statements in this type of texts.

Despite the limitations described, this work contributes to the awareness of the difficulties presented by students when defending their arguments in writing. Knowing how to use arguments to support a point of view will be very useful in other subjects and areas of life, such as – undoubtedly - their teaching career.