Contrastive Corpus Analysis: Investigating the Use of Opinion Verbs in Italian by Greek Learners

In this paper, we investigate the usage of opinion verbs in Italian as a foreign language and any errors this usage may cause, taking into account intralanguage and interlanguage factors. First of all, we analyze the frequency of opinion verb types, in a Greek corpus of native speakers and we investigate the frequency of the same type of verbs in an Italian Translator Learner Corpus compiled by the same learners. Secondly, we explore to what extent there are errors regarding the use of opinion verbs. The data will show that learners, in some cases, commit errors because of their mother tongue influence while in others they avoid using opinion verbs because of this particular lack of knowledge.


Introduction
In this paper, we will present a study on opinion verbs, aiming to provide useful advice for Italian language teachers in Greece, as our main target is to facilitate the learning procedure of the Italian language for Greek learners. For this reason we will use corpus linguistics methods as they are "the most appropriate resource to use among the wealth of other resources available" in pedagogy (Flowerdew 2009). Based on the previous assumption, we examined a Translator Learner Corpus composed by essays written by speakers of the Greek language.

Previous Work
Many scientists have studied parts of speech and, in particular, verbs using Monolingual or Learner Corpora and over the last few years we can find interesting papers combining corpus analysis and research on verbs of the Italian language. Giordano and Voghera (2002) use data coming from the two greatest available Italian corpora: the Lessico di frequenza dell' Italiano Parlato (LIP), for spoken www.scholink.org/ojs/index.php/grhe Global Research in Higher Education Vol. 3, No. 2, 2020 2 Published by SCHOLINK INC.
After analyzing the frequency and the usage of verb types, they have shown that spoken and written texts present more differences at the lexical level than at the grammatical level. Miecznikowski (2006) used the same spoken corpus of the Italian language (LIP), in order to study modal verbs as far as their semantic and pragmatic usage are concerned and how this varies when there is a different verb conjunction. Moneglia et al. (2012) studied in spoken Italian and English corpora the equivalent meanings of verbs of movement in those languages. The 500 most frequent verbs of movement were investigated in both corpora. The experiment consisted of showing pictures that depict the verbs of movement, in order to retrieve the exact meaning and the correspondence between the two languages. Lenci (2014) discussed some methodological problems arising from the use of corpus data for semantic verb classification and he presented a computational framework to describe the distributional properties of Italian verbs using linguistic data automatically extracted from a large corpus.
At the same time, we have an increasing interest for studies in parallel Corpora. Meyers (2002, pp. 22-23) refers to parallel Corpora as a linguistic resource that "facilitates contrastive analysis of English and other languages, advance development in translation theory and enhance foreign language teaching". Bernardini (2003) reports the construction of CEXI (Corpus of English X Italian), a corpus that involves English and Italian original texts and their translations. This corpus was compiled mainly to teach sociocultural insights, discourse analysis, and foreign language aspects. In addition, the multilingual MeLLANGE LTC (Castagnoli 2006) is a project aiming to exploit translator and parallel Corpora using contrastive analyses and linguistic annotation in all six languages involved (Italian language included).

Background
In Greece, choosing the Italian language as a foreign language is rather common. For most learners though, Italian is the third or fourth foreign language that Greeks can learn by taking, mostly, private lessons. On the other hand, English language learning is compulsory from the first grades of the primary education and French or German language learning is given on an optional basis from the age of eleven (Papaefthymiou-Lytra, 2012). Teachers of the Italian language seem to have noticed several patterns that their students have developed, in order to avoid some of the most demanding phenomena of the Italian language. One of these is the way of expressing opinion and the choice of the opinion verb or cluster. The teachers' impression is that verbs of expressing opinion, e.g., "credo", "penso", etc. have always been a trap. Learners of the Italian language tend not to use them in order to avoid using the subjunctive, which is considered a very difficult conjunction. On the other hand, the nature of some essay topics demands the use of those verbs. This automatically triggers, not only a high frequency of opinion verbs, but also a high frequency of errors. In some cases, learners prefer to use other verbs or use another conjunction instead, rather than use the subjunctive. Still, the syntactic rule an opinion support the subjunctive(Note 1)".

Opinion Verbs in Italian Language
According to most grammar reference books there is a vast quantity of verbs and expressions that can be used for expressing opinion such as: credere, pensare, ritenere, supporre, avere l'impressione, ipotizzare, immaginare, non sapere, non essere sicuro (Note 2) and with these verbs the learner is obliged to use the subjunctive in the secondary phrase that follows. However, there are also some structures that can be used for expressing personal opinion without being followed by the subjunctive, e.g., secondo me. Teachers have a certain stereotype regarding learners: They do not use the subjunctive correctly or they avoid using verbs that are related to the subjunctive. However, such a preconception is not based on evidence as, for example, in the case of Spanish learners of the Italian language (Ceruti, 2009). Furthermore, one can ask if the same strategy is followed by advanced learners of Italian language or trainee translators.

Method
As Rawoens (2010) underlined "Multilingual Corpora in cross linguistic research can be used as a tool for verification of hypothesis". Specifically, we test the widely accepted belief that Greek learners of Italian language produce errors when using opinion verbs or they systematically avoid them. To explore this hypothesis, we will examine a subsample of the data from the Multilingual Translator Learner Corpus (MTLC) (Florou, 2019). The translations used were collected from students of the  The next step was to focus on opinion verbs. First of all, we choose to tag all texts for their parts of speech in order to facilitate the retrieval of verb frequency. The tool, which provides the tagged texts, was TagAnt. Soon after using the Antconc tool (Anthony, 2004) we created two frequency lists of all verbs, one from the Greek corpus and one from the Italian translator corpus and in the same way we spotted the clusters that introduced opinion secondary sentences. We detected three opinion verbs in the Greek corpus and two opinion clusters: λοκίδφ, πηζηεύφ, ζεφρώ (Note 3) and ζύκθφλα κε ηελ άπουή κοσ, ζύκθφλα κε ηελ γλώκε κοσ (Note 4). Consequently, we pointed out the opinion verbs and the opinion clusters, which were present into the Italian translator corpus: penso, credo, considero, immagino (Note 5) and second me, secondo la mia opinione (Note 6). At first sight the verbs and the phrases seemed equivalent, but to gain more evidence we had to match the frequency of appearance of those elements in each corpus and, secondly, to check the correct use of the opinion verbs in the Translator Learner Corpus.

Results
Analyzing firstly the Greek texts, we found many verbs that express opinion, probably because of the narrative genre of the texts and the nature of the topics. It is reasonable to assume that a topic that demands narration of personal moments, or a topic that requires the personal opinion of the writer, could lead the learner to use opinion verbs.
Although the findings reveal a vast use of "opinion verbs" in both languages, their presence did not always indicate personal opinion. In many cases, we saw the significance of thought or of belief, e.g.,

[I think of him with nostalgia and love]
As a result, we had to exclude those cases and we listed the verbs and the clusters according to their equivalents, as for their meaning as shown in the following table:

Table 2. The Presence of Opinion Verbs and Clusters in Two Corpora
Greek In the above examples, it is obvious that the learners used the indicative of the verb "essere", as if the rule was never mentioned. The correct form of the phrases should be: Credo cheilgiovanesiainfluenzato, for the first one, and: Credevo che la violenza sia qualcosa normale. The fact that this particular verb ("essere") is among the most frequent verbs, makes the error even more extraordinary. Furthermore, the elevated level of the learners does not correspond to this kind of errors.
The increased possibility of producing these errors, justifies the next phenomenon, a considerable number of zero translation/omissions as shown in the table below: In both examples and in the majority of the cases in which we noticed omission, the learner used the opinion verb and not an opinion cluster in his mother tongue, but avoided the equivalent verb in the Italian language. For a learner, this strategy provokes meaningless changes, but for a translator or an advanced learner, reveals hesitance in using opinion verbs in translation.

Discussion
Teachers' intuition and the research hypothesis of this paper are partially confirmed. We have shown that learners and translator trainees hesitate in their use of opinion verbs, because they are directly related to the subjunctive. Thus, learners use certain opinion clusters that demand the indicative in the corresponding phrase, or they commit grammatical errors by using the wrong conjunction. Additionally, we observed that learners may take a third choice of not using an opinion verb at all, risking a translator error.
This particular behavior can be explained by a number of reasons. It is highly possible for learners to avoid expressing their opinion in the translation, because of their lack of confidence in expressing a personal opinion in the foreign language. The mother tongue (in this case Greek) naturally provides more confidence and familiarity for expressing a personal judgment.
Moreover, learner's interlanguage could be another crucial reason. In the Greek language opinion verbs are followed by a secondary phrase that contains a verb in indicative, which may be a pattern that the learners repeat in a foreign language (in this case Italian).
Having the experience of teaching the Italian language and using most of the course books, we can try to explain learners' choice to avoid the subjunctive by referring to another teachers' stereotype; "subjunctive is so difficult to learn that learners will definitely produce errors".
Teachers themselves provide strategies to students, in order to avoid using subjunctive and sometimes, in order to comfort them on their mistakes, teachers underline how difficult is to use subjunctive, even for native speakers. And this is not an assumption; as Della Valle and Patota (2009) discuss in their book, the subjunctive constitutes a difficulty also for Italian native speakers. As a result, some Italian linguists (Serriani, 2011;Frenguelli, 2017) are already speaking, not only of an erroneous use of subjunctive, but also about the possible extinction of this particular conjunction.
Nevertheless, there is need for further research. It seems useful to cross-examine the above observations by using another Translator Learner Corpus of another language (e.g., English).