You are on page 1of 18

Francesca Costa

Università Cattolica del Sacro Cuore


Cristina Mariotti
Università degli Studi di Pavia, Italy

Differences in content presentation and


learning outcomes in English-­medium
All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or applicable copyright law.

instruction (EMI) vs. Italian-­medium


instruction (IMI) contexts

Abstract: This paper will focus on the acquisition and the presentation of content in ICLHE
(Integrating Content and Language in Higher Education) classes. The learning of content is
indeed one of the most debated but least researched topics in ICLHE settings. In order to
investigate it, this initial study takes into consideration two comparable classes in the field
of Economics taught by the same lecturer throughout one academic year. The first group is
represented by an L1 (Italian) medium instruction class, and the second group by an L2 (Eng-
lish) medium instruction class. Students’ (n. 214) outcomes in terms of marks obtained in two
identically comparable written exams are analysed in order to ascertain whether there are any
differences in learning outcomes between the two groups. Data triangulation is obtained by an
interview with the lecturers and by the comparative analysis of the input presentation strate-
gies of the lecturers in the two groups, collected by means of observations and transcriptions
of the lectures. The results show that, based on students’ marks, the two groups do not differ
significantly as regards learning outcomes. Some differences are to be found in the lecturer’s
input as regards the number of interactions, the use of synonyms, the use of paraphrases and
the use of examples in the two groups. Concerning the lecturer’s interviews, no real differences
Copyright 2017. Peter Lang GmbH, Internationaler Verlag der Wissenschaften.

between the way she/he teaches or in the students’ composition of classes are to be noticed.
Keywords: ICLHE; content acquisition; content presentation; English-­medium instruction;
Italian-­medium instruction

1. Introduction
This paper presents the results of a twofold initial study on the comparison of con-
tent acquisition and the comparison of content presentation in English-­medium
instruction versus Italian-­medium instruction contexts in Italy. The study is part
of a wider one, which includes data from a whole university in Northern Italy.
Dalton-­Puffer (2011) claims that content acquisition has been considered one
of the least researched topics in contexts where a particular discipline is being

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 11/16/2021 7:25 AM via HONG KONG UNIV OF SCIENCE & TECHNOLOGY
AN: 1503000 ; Valcke, Jennifer, Wilkinson, Robert.; Integrating Content and Language in Higher Education : Perspectives on Professional Practice
Account: s8413925.main.ehost
188 Francesca Costa, Cristina Mariotti

delivered through an additional language. However, this is an especially crucial


element to consider at university level, where academic content cannot be simpli-
fied if delivered in an additional language.
The choice of Italy is also particularly relevant because EMI courses have now
reached their plateau (Broggini & Costa, forthcoming) from the beginning of
the Bologna process in 1999. In the wake of this, there is at the moment a debate
on whether these types of courses are in fact desirable, and whether the content
acquisition of the students is hindered when delivered in English instead of Italian
(Maraschio & De Martino, 2013).
Together with Dalton-­Puffer’s research agenda, Pérez-­Cañado’s (2012) sugges-
tions in terms of methodological choices have been chosen so that the present
study comprises both qualitative and quantitative elements and takes the shape
of a quasi-­experimental study.
The following research questions were devised:
• Are the two classes (EMI and IMI) comparable? (This first research question
is also a methodological prerequisite).
• Is there a difference in the learning outcomes (in terms of marks) between an
IMI and an EMI class?
• How does the language of instruction (Italian vs English) affect the lecturers’
teaching style?

2. Literature review
In Italy, the recommendations included in the recent law on universities (Gelmini
Law 240/2010) call for increased teacher and student mobility and for the starting
up of teaching or study programmes in a foreign language (Costa & Coleman,
2012). At the same time, though, there are faculty staff who voice their concern
about the Englishization of higher education because they fear that delivering
subject matter content in English might negatively affect the quality of content
learning itself, ultimately leading to its impoverishment. An example of this is
represented by the Politecnico di Milano, where the decision to teach all Engineer-
ing and Architecture second degree courses exclusively in English has been the
object of discussion over the past few years (Coughlan, 2012). Closely linked to
the issue of student academic performance is the question of how changing the
language of instruction affects the teaching style of lecturers, who might feel that
their ability to convey subject matter content effectively is limited by the use of
English and who might try to prevent comprehension problems by simplifying
their input (Dalton-­Puffer, 2011:188).

EBSCOhost - printed on 11/16/2021 7:25 AM via HONG KONG UNIV OF SCIENCE & TECHNOLOGY. All use subject to https://www.ebsco.com/terms-of-use
Differences in content presentation and learning outcomes 189

Despite the centrality of this topic for ICLHE research, not many studies have
compared the discursive features of lectures delivered in an L1 and in English in
ICLHE contexts, and even fewer studies have analysed the differences in students’
performances across the two teaching modalities. Among the reasons for this
paucity, Dafouz, Camacho, and Urquia (2014:225) list the fact that most EMI
researchers are language specialists adopting a language-­oriented focus rather
than a content-­oriented one; that EMI degrees have been only recently imple-
mented in traditionally monolingual European countries; and, as also observed
by Dalton-­Puffer (2011:188), that the lack of standardized tests across disciplines
and educational contexts make quantitative and cross-­national research difficult
to conduct.
To our knowledge, the studies that have been carried out so far in this field have
mainly focused on primary (Infante, 2009 in Italy; Jäppinen, 2005 in Finland; van
de Craen, Ceuleers, & Mondt, 2007 in Belgium) and secondary levels of educa-
tion (Vollmer, Heine, Troschke, Coetzee, & Küttel, 2006 in Germany; Admiraal,
Westhoff, & De Bot, 2006 in The Netherlands; Badertscher & Bieri, 2009 in Swit-
zerland; Madrid, 2011 in Spain) and have concerned several disciplines, including
art, geography, history and mathematics. Overall, these studies point to the fact
that learning content subjects in a foreign language does not affect the students’
academic performance and show that in English-­medium instruction contexts
students achieve similar results compared to those of their L1-taught counter-
parts. Tertiary education, instead, has been investigated by Dafouz, Camacho, and
Urquia (2014), who have analysed the academic outcomes of first-­year Spanish
students enrolled in three types of BA degree (accounting, finance and history) at
the Complutense University of Madrid and have found that a change in the lan-
guage of instruction did not affect the learning outcomes of the students involved
in the study. Moreover, Hernandez-­Nanclares and Jimenez-­Munoz (2015) carried
out a two-­year research project in the Faculty of Economics at the University of
Oviedo, Spain, and compared the test results of the English-­taught class with those
of the Spanish-­taught one. They found that the two groups had similar grades,
but the English-­taught group did not perform as well as the Spanish-­taught one
as far as the higher band is concerned.
Among the scholars who have investigated the relationship between language
change and teaching style, Vinke (1995) researched the topic extensively by ad-
ministering a questionnaire to 131 lecturers at a technical university in the Neth-
erlands, accompanied by a follow-­up study where she recorded 16 engineering
lectures in English and Dutch. She took into consideration the amount of lecturer
verbalization (words per minute), the presentation of information with particular

EBSCOhost - printed on 11/16/2021 7:25 AM via HONG KONG UNIV OF SCIENCE & TECHNOLOGY. All use subject to https://www.ebsco.com/terms-of-use
190 Francesca Costa, Cristina Mariotti

reference to lecture structuring (signposting), redundancy (summarizing, re-­


stating, providing examples), and the presence of interaction and compensation
(i.e. behaviours displayed by lecturers to overcome limitations in their oral English
proficiency). The results show that there are some differences linked to the lectur-
ers’ individual teaching profiles, the group size and the type of course (graduate/
postgraduate). In both contexts, lecturers held turns almost exclusively, even if
student-­lecturer interactions appear to occur more frequently when lecturers
teach in English. Moreover, Dutch lecturers are less redundant in English and they
tend to use structuring techniques more frequently in English than in Dutch. For
most of the lecturers the speech rate is reduced by an average of 17 % when they
speak in English. The lecturers say they hardly notice any difference in teaching in
English or Dutch, but they report an increase in preparation time needed for EMI.
It should be noted that this data refers to lecturers who are very experienced and
teach in English on a daily basis, so findings may not be generalized to contexts
with less experienced teachers.
In the same technical university in the Netherlands, Klaassen (2001) studied
the relationship between lecture intelligibility and the language competence and
pedagogical approach of the lecturers. Lectures were video-­recorded and rated for
comprehensibility and student-­centredness, and the findings show that student-­
centered lecturing is a much more important factor in the success of a lecture
than the lecturer’s language competence, which in the case of this university is
represented by a C1 threshold level.
Crawford Camiciottoli (2005) compared the changes that a native-­speaking
lecturer introduced in a lecture on Japanese economy addressed to an Italian
audience versus the original delivery to a UK audience. The findings point at a
slower speech rate, more redundancies, a higher number of interpersonal features
(elicitation, personal pronouns) and references to the local culture of the speaker
as the characteristics of the intercultural lecture.
Thøgersen and Airey (2011) analysed five science lectures (three in Danish and
two in English) given by the same lecturer, finding that the lecturer took longer
to present the same subject matter in English, speaking more slowly, using more
repetitions in L2 and more formal language, similar to written, textbook style.
Finally, Arkin and Osam (2015) conducted a study on a lecturer teaching the
same content in Turkish and in English to undergraduate university students
studying Human Resources Management at an English-­medium university in
Turkey. Findings show that on average in the English-­medium lectures it took the
lecturer 11 % more time to present the same content in English, and that he did
not have time to provide extra examples, to cover all the intended material and to

EBSCOhost - printed on 11/16/2021 7:25 AM via HONG KONG UNIV OF SCIENCE & TECHNOLOGY. All use subject to https://www.ebsco.com/terms-of-use
Differences in content presentation and learning outcomes 191

allow time for student questions and discussion. Moreover, the lecturer tended to
use self-­repetitions most of the time in a conscious effort to help comprehension.
Arkin and Osam also conducted student interviews where the subjects reported
limited comprehension and even misunderstanding of disciplinary content. Thus,
the authors claim that the observed EMI setting seems to be adversely affecting
the task of content learning and suggest a move from EMI to CLIL for university
settings, where a language teacher addresses the academic language needs of the
learners and the content instructor employs strategies to help students cope with
the content and language requirements of the course (2015:195)

3. Methodology
The methodology of this study comprises a triangulation of data by means of:
1) Interview with the lecturers, 2) Comparison of content acquisition in IMI and
EMI expressed by means of marks obtained by students in their final test, and
3) Discursive differences in EMI and IMI (Table 1). In order to draw conclusions
on the effect of language change on the teaching style of lecturers and the learn-
ing outcomes of students, the only dependent variable should be represented by
the change in the language of instruction itself; all the remaining factors should
be kept as similar as possible, with the exception of the students, who obviously
cannot be the same. Even if it is virtually impossible to keep all the independent
variables under control, as this would imply making sure that students do not
receive input from other sources (including exposure to input outside the class-
room), to satisfy the basic methodological requirements of the present study it was
decided that the lecturers should teach the same content in two parallel courses
to similarly sized, same-­level groups, possibly in the same university.

3.1 Sample
As regards the lecturers, the sample was chosen by means of a criterion sampling;
therefore the lecturers had to teach the same course to the same level of students
both in Italian and English, they had to be Italian native speakers and they had to
be from the faculty of Economics in a university in the North of Italy (where the
majority of these courses are carried out). Moreover, it was very important that
the students in the EMI and IMI classes would be tested with exactly the same
exam paper. After having looked at the websites of various universities where Eco-
nomics was taught, an email was sent to professors who taught in both languages.
This initial study started with two lecturers: one female professor of Economics
(12 years of experience; 8 years with parallel groups) and one male professor of

EBSCOhost - printed on 11/16/2021 7:25 AM via HONG KONG UNIV OF SCIENCE & TECHNOLOGY. All use subject to https://www.ebsco.com/terms-of-use
192 Francesca Costa, Cristina Mariotti

International Relations (15 years of experience; 2 years with parallel groups). The


two lecturers come from two different universities.
As regards the sampling of the students, the relative sample included 131 EMI
students and 83 IMI students. The certified proficiency level of the English class
was at least C1, as per enrolment requirements of the Faculty. They were all in
the same year of study, on the same three-­year course (Master’s course/Laurea
Magistrale). As the study focused on their anonymous marks in the final exam,
no data on their sociolinguistic background was investigated.

3.2 Instruments
In order to determine whether the two classes (IMI and EMI) were comparable,
two interviews with the teachers concerned were held. These were carried out in
Italian and translated by one of the researchers with the other one acting as an au-
dit. For the analysis of the interviews a content analysis was carried out (Gillham,
2000); that is to say, only comments relevant to this research were transcribed.
The interview with the professor had the objective of understanding whether the
two classes were comparable and to understand the teachers’ opinions on EMI
(Table 1). Although the choice of using the interviews as a methodological prere­
quisite has evident flaws, it was seen as the most feasible and non-­obstructive one.
Since the interviews confirmed that the two classes were comparable, the lecturers
were asked to let the researchers observe and record their classes.

Table 1. Purposes of the study and instruments used.

IMI/EMI
Interview Understand whether the two groups were comparable/view
lecturers’ opinions on EMI or IMI
Student t test Compare exams marks in the two types of teaching
Discourse Analysis Compare discursive differences of the lectures when teaching
to the two different groups

One of the lecturers agreed to having two of her lectures on anti-­trust law (one
in the Italian and the other in the English-­taught class) recorded, transcribed and
analysed. The English lecture lasted 85 minutes and the Italian one 79 minutes;
therefore, it was decided to take into consideration only the speech produced
during the first 79 minutes of the English lecture to make them comparable. It
was not possible to record the second lecturer because he had already finished his
term lectures. The lecture was analysed both from a qualitative and a quantitative

EBSCOhost - printed on 11/16/2021 7:25 AM via HONG KONG UNIV OF SCIENCE & TECHNOLOGY. All use subject to https://www.ebsco.com/terms-of-use
Differences in content presentation and learning outcomes 193

point of view, whereby textual and pragmatic features were identified. The speech
produced by the lecturer was coded into categories which reflect how content in-
formation was made accessible to students with reference to: lecture structuring,
redundancy, degree of interactivity and speech rate. The occurrences were then
counted and analysed using descriptive statistics.

4. Interview analysis
In general, it should be noted that the female teacher speaks more. When asked
directly, neither of the two teachers noted any differences between the two groups.
However, talk of a kind of self-­motivation amongst the EMI classes (the EMI
groups as being more stimulated) is present in their discourse. Nonetheless over
time it appears that this notion is declining.
The male teacher has been teaching for 15 years, and for 10 in English; he has
only had the parallel EMI and IMI groups two years. He judged the overall experi-
ence as positive. The average number of students attending the Italian classes was
15, and 25 for the English. The degree is a Master’s degree. When asked about the
differences between Italian and English, he stated that:
The class in Italian is more formal, more full-­on, and more rigid. Conversely in an inter-
national class students get more involved than in the Italian classes (with more seminars
and question and answer sessions).

He also asserts that he never noticed substantial differences in either of the two
groups in terms of grades. However, when asked if the two groups are compara-
ble, he said:
The group taught in English was much more homogenous in terms of learning outcomes
and previous knowledge (even if they came from other universities).

He maintained, however, that:


The method of assessment is more severe and demanding for the class taught in English
(as it has a more focused and international curriculum), than the class taught in Italian,
which follows a less specific curriculum on the study of international relations.

The female professor has been teaching for around 12 years and for 8 years she
has been teaching her subject to two parallel groups of students: one in English
and one in Italian.
When asked whether there are differences between the two classes, she re-
marked:
In my opinion it is a complicated experience, in the sense that it forces the students to
abandon their usual method of self-­expression and actually revise the content they are

EBSCOhost - printed on 11/16/2021 7:25 AM via HONG KONG UNIV OF SCIENCE & TECHNOLOGY. All use subject to https://www.ebsco.com/terms-of-use
194 Francesca Costa, Cristina Mariotti

learning. Moreover, it is difficult to fully replicate what you would do in an Italian class
in English, even if your English was excellent. It is something that requires new channels
of expression.

The teacher also commented that it is not in fact possible to teach two identical
courses, especially when taking cultural differences into account, in terms of the
content that is taught. In regard to these differences, the teacher noted that it took
an initial period of time to adjust to having a slightly different teaching identity.
She did not feel entirely herself when teaching in English; therefore she had to
learn the language better in order to get rid of this first period of adjustment, thus
making way for her new identity as an EMI teacher.
No I do not think so. I believe I needed to find ways to be myself in both languages. I tend
to be myself, but in my first few years teaching in English for instance, I was not able to
joke around with the students due to me not being fluent, meaning I came across as more
severe than I am in reality, simply because I was not able to convey the same nuances and
my language came across harsher. There is a problem here, in that it is difficult sometimes
to represent your personality in a language, which during my life has not represented
who I am. The solution, I found over time, was not to change myself, but to learn the
language better in order to represent better my true personality.

The teacher also commented on the importance of having an adapted language


for the course.
I have not changed my stance on thinking that it is essential for a teacher to be able to
speak good English, especially in terms of establishing authority in front of students.
When English students become more able, it is important that your English fits their level,
otherwise the students will not recognize or respect the role of the teacher. They begin
to enquire and say “if I know what I’m doing and you do not, there is not a lot of use in
you being here.” Thus, teachers also have to make an investment on their language skills.

When asked solely about the linguistic differences between doing the lesson in
Italian or in English, she said:
In Italian I use the past historic and imperfect subjunctive tenses a lot but because I am
from the South I also speak in with hypothetical and conditional tenses and phrases,
something I am proud of. In English I struggle to manage to construct complete phrases
in the same way, I try in any case, but I am aware of myself making grammatical mistakes
and thus tend to speak in a more simple manner, more in the present and future than
the imperfect subjunctive.

As a result, the professor had the impression of simplifying her speech when
speaking in English; however, this is also down to her cultural background, which
in terms of Italian speaking demands much verbosity and eloquence. As regards
the sample of students the professor affirmed:

EBSCOhost - printed on 11/16/2021 7:25 AM via HONG KONG UNIV OF SCIENCE & TECHNOLOGY. All use subject to https://www.ebsco.com/terms-of-use
Differences in content presentation and learning outcomes 195

Most often there are usually around 80 students on average attending lectures. Moreover,
this is a graduate programme with a male prevalence of 55 %, with 45 % women, due to
the fact that it is a course with a level of high quantitative content, comprising 2 years of
lots of mathematics and econometric statistics, which tends to attract more males than
females.

When asked if she noticed any differences between the classes in terms of interac-
tion, she commented:
No there was not really any differences in a cross-­group sense, however the behaviour
in the English class changed over time (across-­time sense). They interacted less at first
because they were a little shy about their English, but within 2 or 3 years we had students
that spoke an excellent level of English, which was also decisively better than mine.
There are many students with whom it is obvious that they had studied English from a
young age, and would therefore interact freely. However I would say that as the class is
linguistically a fairly controlled environment, in that I understand them well as they are
discussing topics that I know better than them, therefore I understand the meaning they
are trying to convey when they speak.

Conversely, however, as regards the general differences between the two groups
she remarked:
Over the past few years I have always had the feeling that the Italian classes were less
talented than the English classes, but over this year I have seen that the students in the
English classes have been in fact more prepared, more brilliant, more ambitious and more
determined to set out and conquer the world. I had the sense that they came from more
wealthy classes and had had more sophisticated experiences in their life, for instance
studying and travelling regularly abroad. […] I no longer have this impression. This year
the Italian class are just as good as those in the English class.

Therefore when asked directly if the two classes are comparable she said: definitely
comparable. When asked if she has the impression that the exams tend to go in
different ways because of the language differences, she remarked:
No, there were reasonably similar errors, I did not notice great differences between the
two groups. We cannot say that one class studied more than the other […] and we cannot
say that one did better than the other. The only differences […] are that in class the respon-
siveness and maturity levels expressed seemed to be a little higher in the English class.
Less fearful and more determined. This is more a personal account than professional.

5. Statistical analysis of exam marks


For the data analysis a t-­test was used to determine how much the variation in
scores between the two groups derived from either the difference between the
groups (signal) or the differences within the group (noise). First of all, it was

EBSCOhost - printed on 11/16/2021 7:25 AM via HONG KONG UNIV OF SCIENCE & TECHNOLOGY. All use subject to https://www.ebsco.com/terms-of-use
196 Francesca Costa, Cristina Mariotti

decided to calculate the average of the two groups. In Italy the range of marks at
the university level is from 18 (pass mark) to a maximum of 30.

Table 2. Means of the exam marks according to disciplines.

Economics International Relations Total


Italian mean 26.58 27.08 26.69
English mean 27.53 27.36 27.50

As can be seen from the means, the marks in the English class prove to be slightly
higher, for both subjects. The results demonstrate that in no case is the difference
of marks between the EMI classes and the IMI classes significant. Therefore, the
two groups did not differ in learning content when using as a parameter the marks
given in the examinations.

6. Discourse analysis of the transcriptions of the lectures


In the present analysis it was decided to look for occurrences of discursive fea-
tures that enhance input comprehensibility and lecturer-­students interactivity
since, as the literature shows, they seem to be central in EMI research. Moreover,
in their analysis of lecture features, Flowerdew and Miller (1997:38) stress the
importance of making information clearly accessible to students in several ways
and underline the relevance of redundancy and explicitness for academic didac-
tic spoken discourse. This can be achieved by establishing clearly the theme of
the lecture right from the beginning, using a narrative thread to hold the lecture
together, asking rhetorical questions to signal lecture structure, using micro-­level
discourse markers such as ‘and’, ‘so’, ‘now’ and ‘okay’, and pauses to signal tone
groups and propositional boundaries, and signposting the correlations between
concepts through the use of macro-­markers, for example, “Okay, let’s get started”,
or “now here / we’ll put up our last slide / and come to the conclusions” (see also
Chaudron & Richards (1986) and Nattinger & DeCarrico (1992) on the relation-
ship between macro-­markers and lecture comprehension).
Paraphrases also represent a key strategy for making input easier to process,
and therefore more comprehensible, while keeping it semantically rich both in for-
mal language learning contexts (Met, 1994 in ESL contexts and Dafouz-­Milne &
Llinares-­Garcia, 2008 in CLIL contexts) and in conversation between native and
non-­native speakers (Long, 1996). Finally, display questions (questions where
the answers are already known by the questioner), and comprehension checks
(interactional moves by which one speaker attempts to determine whether the

EBSCOhost - printed on 11/16/2021 7:25 AM via HONG KONG UNIV OF SCIENCE & TECHNOLOGY. All use subject to https://www.ebsco.com/terms-of-use
Differences in content presentation and learning outcomes 197

other speaker has understood a preceding message) are frequently used in di-
dactic discourse in general and in CLIL contexts in particular (Dafouz-­Milne &
Sanchez-­Garcia, 2013) as they promote student participation in discourse, thus
potentially improving both content and language learning, and enable lecturers
to verify student comprehension.
Drawing on these considerations, it was decided to look for occurrences of
discursive features that may help learners access content effectively. The following
categories were identified: micro and macro discourse markers, rhetorical ques-
tions and explicit signalling of particularly relevant points for lecture structuring;
paraphrases, reformulations, repetitions and examples for providing redundancy;
speech turns, number of comprehension checks, display and referential questions
for interactivity; speech rate for clarity.
Overall, in the observed lectures structuring strategies appear more frequently
in the EMI class where we can observe 19 occurrences of macro and 64 of micro
discourse markers versus 10 and 17 respectively in the IMI class. Rhetorical ques-
tions, i.e. questions by which the lecturer makes a point rather than elicit answers
from students, are slightly fewer in the EMI (22 occurrences) than in the IMI (27
occurrences). An example is provided in (1)1:
(1)  now what does the dominant firm do, it says look since I’m dominant I have many
many market ehm a large market share

At times, the lecturer also brought some particular questions to the attention of
students very explicitly, as in the following excerpts:
(2)  listen to me because this is important
(3)  what I want you to remember is that you are totally fine in assuming this if you have
clear in your minds that consumer welfare, the protection of consumer welfare is the goal
(4)  [literally shouting] that’s competition guys!

Nevertheless, she did so in a more balanced way across the two teaching modali-
ties, with 4 occurrences in the EMI class and 6 in the IMI class. Lecture structuring
strategies tend to be more explicit in the EMI setting, where they occur almost
twice as often (EMI 109 vs. IMI 60).

1 Transcription conventions:
[comment of the observer]
, short pause (up to 3 seconds)
. longer pause (more than 3 seconds)
! increasing volume
? rising intonation

EBSCOhost - printed on 11/16/2021 7:25 AM via HONG KONG UNIV OF SCIENCE & TECHNOLOGY. All use subject to https://www.ebsco.com/terms-of-use
198 Francesca Costa, Cristina Mariotti

The degree of redundancy was calculated based on verbatim repetitions (22


in EMI vs 11 in IMI), syntactic reformulations (EMI 3 vs IMI 0) and semantic
reformulations or paraphrases (EMI 87 vs 63 IMI). In (5) an example of syntactic
reformulation is provided, whereas (6) and (7) show instances of paraphrases
from the corpus:
(5)  its marginal cost, the marginal cost of the dominant firm
(6)  if you did the other way round, if you strike down this reduction of price
(7)  we cannot say how long that monopoly will last, we cannot regulate the duration
of monopolies

Explicit definitions, signaled by introductory clauses such as “that means”, were


counted as paraphrases, for instance:
(8)  intellectual property rights, are legislative administrative barriers to entry. that means
that in order to reproduce the technical improvement innovation the creation which is
covered by intellectual property right you must have the consent of the holder

It is interesting to observe that in the EMI class the lecturer used paraphrases to
reword student utterances twice, as in (9):
(9)  so, so you say, that this change in market price this change in the price of the domi-
nant firm who is the leader so if it changes price the market price changes as well

This may show awareness on the part of the lecturer that paraphrasing can be an
effective tool for making input at the same time clearer and more abundant, so
much so that she deploys this strategy extensively by paraphrasing also student
speech in addition to her own utterances.
Finally, giving examples was decidedly more frequent in the EMI class, where
they were used 8 times versus only once in the IMI class:
(10)  It prevents the reproduction of the dominant technology or the the technology they
cover, if that technology is dominant then it is much more difficult to enter the market
unless you invent around that technology, do you get it? alright, so, this was just an ex-
ample, that I gave you to, to fix in your mind this reasoning which is the reasoning that
economists do in order to establish whether a conduct is exclusionary and anticompetitive

Overall, the data show that the lecturer tends to adopt several strategies to make
her input semantically clear and that she does so 51 % more frequently in the
EMI than in the IMI class.
The degree of interactivity was calculated based on speech turns and on the
number of questions asked by the lecturer. The following types of questions were
coded and counted:

EBSCOhost - printed on 11/16/2021 7:25 AM via HONG KONG UNIV OF SCIENCE & TECHNOLOGY. All use subject to https://www.ebsco.com/terms-of-use
Differences in content presentation and learning outcomes 199

– comprehension checks (32 EMI vs 35 IMI), which have the function of making
sure that what was said has been understood by the students,
(11)  do you get the meaning of what I’m saying?

– display questions (39 EMI vs 23 IMI), where the lecturer elicits answers from
students to check on their comprehension and learning processes, sometimes
very openly as in the following example:
(12)  so at this point would you strike down the conduct saying that it is an anti-­trust
violation yes or no? . you must have an opinion guys!

It was found that the English lecture was more interactive both judging by the
number of student turns (74 in the EMI vs 60 in the IMI class amounting to a
23 % difference) and the number of display questions uttered by the lecturer
(EMI 39 vs. IMI 23 amounting to a 70 % difference). It should be noted that in
both classes students generally sounded very confident regardless of the lan-
guage of instruction, and that they started negotiation sequences by asking
questions with a similar frequency (17 in the EMI and 18 in the IMI class). In
both classes they even interrupted the lecturer’s turn three times to negotiate
for meaning by asking for clarifications or more information on the case studies
presented during the lecture, thus showing that the change in the language of
instruction did not inhibit their active participation in class. The speech rate
of the lecturer was around 12 % slower in English with 103 words per minute
versus 115 in Italian.
Overall, the results are in keeping with what was found in previous studies
as far as the speech rate is concerned (see for instance Vinke, 1995 and Vinke,
Snippe, & Jochems, 1998), but they do not confirm Vinke’s findings (1995) about
redundancy and lecture structuring since the lecturer observed in the present
study tends to be more redundant in English and extensively guides the students
throughout the lecture by means of signposting. In the English-­taught class the
NNS lecturer tends to ask more display questions (as was also observed by Ar-
kin & Osam, 2015), and students tend to react to questions more than in the
Italian-­taught one. It is interesting to observe that both contexts show a similar
number of student self-­selections, hinting at the fact that in these classrooms
students who need clarifications negotiate for meaning even if they have to use a
foreign language to do so.

EBSCOhost - printed on 11/16/2021 7:25 AM via HONG KONG UNIV OF SCIENCE & TECHNOLOGY. All use subject to https://www.ebsco.com/terms-of-use
200 Francesca Costa, Cristina Mariotti

7. Discussion and conclusions


This paper has sought to analyse the differences in content acquisition and pres-
entation in IMI and EMI lectures. Starting from our research questions this study
has provided the following results.
As for the comparability of the two different groups (IMI and EMI) the lectur-
ers state that they are comparable. At times, though, they highlight the possibility
of a kind of self-­selection in the EMI group compared to the IMI group, although
this is still hypothetical and would need further empirical data.
As regards content acquisition measured by means of final exam marks, no evi-
dence was found of significant differences, even though from a strictly mathemati-
cal mean point of view the EMI groups performs slightly better. Although results
are still in the initial phase and more data are needed in order to generalize find-
ings, this study finds no support for claims that EMI hinders content acquisition.
As for the differences in teaching style, the discursive data show that in the EMI
group the lecturer tended to produce more signposting and structuring speech to
help the students orient themselves throughout the lecture; to be more redundant
and semantically clear by means of repetitions, paraphrases and examples; and to
adopt a slower speech rate. These strategies can hint at a greater awareness on the
part of the lecturer (which is also confirmed in the interview) of the fact that the
change in the language of instruction requires changes also in her teaching style,
and that these changes are geared towards increasing input comprehensibility
and helping students process the content by making the language more explicit.
Previous studies, such as Arkin and Osam’s (2015: 193), argue that this might
negatively affect the quality of subject matter learning. Nevertheless, in the present
study the lecturer reports covering the same amount of content in the two courses;
moreover, the discursive data show that in the EMI she had time to provide more
examples than in the IMI and to allocate the same time for student questions.
Considering that the statistical data analysis shows that the two groups did not
differ in learning content when using the final examination marks obtained by
students as a parameter, it may be safely concluded that in the observed setting
the change in language of instruction did not slow down or compromise the
content learning process.
As far as the degree of interactivity is concerned, in the EMI class the lecturer
tended to ask more display questions, to which students reacted more often than
their IMI counterparts. It might be concluded that the EMI class shows a higher
degree of interactivity because it is more likely to include Italian students or for-
eign students who have already had exposure to English-­taught classes abroad,
and therefore are possibly more used to an interactive lecturing style based on

EBSCOhost - printed on 11/16/2021 7:25 AM via HONG KONG UNIV OF SCIENCE & TECHNOLOGY. All use subject to https://www.ebsco.com/terms-of-use
Differences in content presentation and learning outcomes 201

problem solving and critical thinking, which is often encouraged in Anglophone


teaching contexts.
Summing up, the discursive differences (i.e. greater redundancy, higher degree
of interactivity, and slower speech rate) in the speech produced by the lecturer
across the EMI and the IMI classes, and observed also in some of the studies
previously carried out in EMI settings in Europe, are candidates to be consid-
ered as distinctive of EMI contexts. Nevertheless, further, possibly longitudinal
studies need to be carried out to gather more comparative data on how the use
of the English language affects the teaching style of lecturers and to determine
whether the change in the language of instruction has an impact on the students’
learning outcomes.
This was an initial study that allowed the researchers to detect its limitations,
which will be dealt with in the wider study with a bigger sample. The fact that
only one lecturer could be recorded was purely linked to the situational context
in the sense that the second lecturer would have agreed to be observed had he
not finished his term lectures. Moreover, the decision to use lecturers’ interviews
as a methodological prerequisite in order to test whether the EMI and the IMI
groups of students were comparable was a compulsory one since neither students
nor lecturers would agree to a pre-­test and post-­test method of analysis.

References
Admiraal, W., Westhoff, G., & de Bot, K. (2006). Evaluation of bilingual second-
ary education in the Netherlands: Students’ language proficiency in English.
Educational Research and Evaluation, 12, 75–93.
Arkin, E. & Osam, N. (2015). English-­medium higher education: A case study in
a Turkish university context. In S. Dimova, A.K. Hultgren, & C. Jensen, (Eds.),
English-­medium instruction in European higher education (pp. 177–199). Bos-
ton/Berlin: De Gruyter Mouton.
Badertscher, H. & Bieri, T. (2009). Wissenserwerb im content and language inte-
grated learning: Empirische Befunde und Interpretationen [Knowledge acquisi-
tion in content-­and-­language-­integrated learning: Empirical evidence and inter-
pretations]. Bern/Stuttgart/Wien: Haupt.
Broggini, S. & Costa, F. (forthcoming). A survey of English-­medium instruction
in Italian higher education. A longitudinal perspective from 2012 to 2015.
Chaudron, C. & Richards, J. (1986). The effect of discourse markers on the com-
prehension of lectures. Applied Linguistics, 7(2), 113–125.

EBSCOhost - printed on 11/16/2021 7:25 AM via HONG KONG UNIV OF SCIENCE & TECHNOLOGY. All use subject to https://www.ebsco.com/terms-of-use
202 Francesca Costa, Cristina Mariotti

Costa, F. & Coleman, J. (2012). A survey of English-­medium instruction in Italian


higher education. International Journal of Bilingual Education and Bilingual-
ism, 15(4), 1–17.
Coughlan, S.  (2012, May 16). Italian university switches to English. BBC
News. Retrieved 28th September 2016 from: http://www.bbc.co.uk/news/
business-17958520
Crawford Camiciottoli, B. (2005). Adjusting a business lecture for an international
audience: A case study. English for Specific Purposes, 24(2), 183–199.
Dafouz, E., Camacho, M., & Urquia, E. (2014). ‘Surely they can’t do as well’: A
comparison of business students’ academic performance in English-­medium
and Spanish-­as-­first-­language medium programmes. Language and Education,
28(3), 223–236.
Dafouz Milne, E. & Llinares García, A. (2008). The role of repetition in CLIL
teacher discourse: A comparative study at secondary and tertiary levels. Inter-
national CLIL Research Journal, 1(1), 50–59.
Dafouz Milne, E., & Sánchez García, D. (2013). ‘Does everybody understand?’
Teacher questions across disciplines in English-­mediated university lectures:
An exploratory study. Language Value, 5(1), 129–151.
Dalton-­Puffer, C. (2011). Content and language integrated learning: From practice
to principles? Annual Review of Applied Linguistics, 31, 182–204.
Flowerdew, J. & Miller, L. (1997). The teaching of academic listening comprehen-
sion and the question of authenticity. English for Specific Purposes, 16(1), 27–46.
Gillham, B. (2000). Case study research methods. London: Continuum.
Hernandez-­Nanclares, N. & Jimenez-­Munoz, A. (2015). English as a medium of
instruction: Evidence for language and content targets in bilingual education
in economics. International Journal of Bilingual Education and Bilingualism,
[online] 1–14. DOI: 10.1080/13670050.2015.1125847
Infante, D. (2010). Il Content and Language Integrated Learning (CLIL) in Italia,
Modelli didattici e sperimentazioni nella scuola primaria. Roma: Editrice Nuova
Cultura.
Jäppinen, A.K. (2005). Thinking and content learning of mathematics and science
as cognitional development in Content and Language Integrated Learning
(CLIL): Teaching through a foreign language in Finland. Language and Educa-
tion, 19, 148–169.
Klaassen, R.G. (2001). The international university curriculum: Challenges in
English-­medium engineering education. Delft: Technische Universitat Delft.
Retrieved 28th September 2016 from: http://repository.tudelft.nl/view/ir/
uuid%3Adea78484-b8c2-40d0-9677-6a508878e3d9/.

EBSCOhost - printed on 11/16/2021 7:25 AM via HONG KONG UNIV OF SCIENCE & TECHNOLOGY. All use subject to https://www.ebsco.com/terms-of-use
Differences in content presentation and learning outcomes 203

Long, M. (1996). The role of the linguistic environment in second language ac-
quisition. In W.C. Ritchie, & T.K. Bhatia (Eds.), Handbook of second language
acquisition (pp. 413–468). San Diego: Academic Press.
Madrid, D. (2011). Monolingual and bilingual students’ competence in social
sciences. In D.  Madrid,  & S.  Hughes (Eds.), Studies in bilingual education
(pp. 195–221). Frankfurt: Peter Lang.
Maraschio, N. & De Martino, D. (2013). Fuori l’italiano dall’università? Inglese,
internazionalizzazione, politica linguistica. Roma/Bari: Laterza.
Met, M. (1994). Teaching content through a second language. In F. Genesee (Ed.),
Educating second language children (pp. 159–182). Cambridge: Cambridge Uni-
versity Press, Cambridge.
Nattinger, J.R. & DeCarrico, J. (1992). Lexical phrases and language teaching. Ox-
ford: Oxford University Press.
Pérez-­Cañado, M.L. (2012). CLIL research in Europe: Past, present and future.
International Journal of Bilingual Education and Bilingualism, 15(3), 315–341.
Thøgersen, J. & Airey, J. (2011). Lecturing undergraduate science in Danish and in
English: A comparison of speaking rate and rhetorical style. English for Specific
Purposes, 30(3), 209–221.
van de Craen, P., Ceuleers, E., & Mondt, K. (2007). Cognitive development and
bilingualism in primary schools: Teaching maths in a CLIL environment. In
D. Marsh & D. Wolff (Eds.), Diverse contexts–­Converging goals: CLIL in Europe
(pp. 185–200). Frankfurt: Peter Lang.
Vinke, A.A. (1995). English as the medium of instruction in Dutch engineering
education. Doctoral thesis. Delft: Technische Universiteit Delft.
Vinke, A.A., Snippe J., & Jochems W. (1998). English-­medium content courses in
non-­English higher education: A study of lecturer experiences and teaching
behaviours, Teaching in Higher Education, 3(3), 383–394.
Vollmer, H.J., Heine, L., Troschke, R., Coetzee, D., & Küttel, V. (2006). Subject
specific competence and language use of CLIL learners: The case of geography
in grade 10 of secondary schools in Germany. Paper presented at the ESSE 8
Conference, London, UK.

Francesca Costa is currently a university lecturer in English Linguistics at Uni-


versità Cattolica del Sacro Cuore in Milan. Since 2001 she has been involved in
teaching and research at all levels of education on Content and Language Inte-
grated Learning (CLIL) and English-­medium Instruction (EMI). She has pub-
lished several articles on these subjects in both national and international journals.
francesca.costa@unicatt.it

EBSCOhost - printed on 11/16/2021 7:25 AM via HONG KONG UNIV OF SCIENCE & TECHNOLOGY. All use subject to https://www.ebsco.com/terms-of-use
204 Francesca Costa, Cristina Mariotti

Cristina Mariotti is Associate Professor of English in the Department of Political


and Social Sciences at Pavia University, Italy. Her research is focused on English
as a medium of instruction. She is the author of the volume Interaction Strategies
in English-­medium Instruction (Franco Angeli, 2007), and she has co-­edited the
volume Subtitles and Language Learning (Peter Lang, 2014). cristina.mariotti@
unipv.it

EBSCOhost - printed on 11/16/2021 7:25 AM via HONG KONG UNIV OF SCIENCE & TECHNOLOGY. All use subject to https://www.ebsco.com/terms-of-use

You might also like