Designing a Corpus Workbook
for Students of Czech as a Foreign Language
Adrian Jan Zasina
1 INTRODUCTION
Language corpora are increasingly utilised in language learning around the world.
The interest in using corpus methods for learning Czech appeared at the beginning of
the 21st
century. In 2005, a team of scientists from the Institute of the Czech National
Corpus (Čermák et al., 2005) published the first study guide on how to use the Czech
corpus and recommended it as a supplementary source for students from primary
and secondary schools and universities. Two years later, Šulc (2007) described in
a short guide for schools how to work step by step with corpus data; he explained the
possible functions and usage of the Czech corpus. Over the next four years, the first
works concerning corpus use in teaching Czech as a foreign language (CFL) started
to appear (Vališová, 2009; Lukšija, 2010; Osolsobě, 2010). Since then, several studies
have appeared on corpora use in CFL teaching (cf. Lukšija, 2012; Vališová, 2012, 2016;
Kočařová, 2013; Konečná & Zasina, 2014; Zasina, 2018b). The last few years have seen
publications aimed at Czech teachers and students of teacher education for primary
and secondary schools; these works (Šmejkalová & Kopřivová, 2019; Šormová et al.,
2019) attempt to incorporate corpus methods developed for teaching native speakers
of Czech. Zasina’s (2019) comprehensive study describes the most prominent problematic
areas of non-native Czech based on a learner corpus and offers a coherent
methodological framework for creating corpus-based exercises.
Despite a great deal of interest in corpus approaches to teaching, no substantial
collection of corpus-based exercises has been created. Teachers are reluctant to use
corpora in the classroom as it is a time-consuming task requiring computer skills and
knowledge of the potential of corpus tools that might not be user-friendly (Bennett,
2010, pp. 206–207; Tribble, 2015, pp. 57–59; Leńko-Szymańska, 2017, p. 217). Thus, it is
especially important to provide teachers with suitable learning materials that they
could easily use. Giving them ready-to-use materials could help them to actively apply
corpora in teaching.
This paper aims to familiarise teachers with the method of data-driven learning,
describe the design of the corpus workbook, and recommend examples of corpusbased
exercises.
2 DATA DRIVEN LEARNING: PROS AND CONS
OF USING CONCORDANCE IN TEACHING
Using a corpus in foreign language learning was termed data-driven learning (DDL) by
Tim Johns (1991). In his seminal work, DDL is defined as an:
OPEN
ACCESS
126STUDIE Z APLIKOVANÉ LINGVISTIKY 2/2022
approach to foreign language learning that takes seriously the notion that the
task of the learner is to “discover” the foreign language, and that the task of the
language teacher is to provide a context in which the learner can develop strategies
for discovery — strategies through which he or she can “learn how to learn”.
(ibid., p. 1)
As proposed by Johns (1991, p. 4), it is an inductive approach to language learning
built on three pillars: identification — classification — generalisation. Students identify
a language problem they would like to solve; then they observe a concordance
with examples to follow language features and classify them; finally, they generalise
rules about how a language phenomenon works. It means that students do not obtain
information, for example, about the function and formation of the accusative case,
but instead must derive it based on concordance observation: following concordance
lines, students identify that the accusative case has its objective function and specified
endings for each grammatical gender.
The usage of concordance in teaching is often underestimated and criticised. According
to Boulton and Cobb (2017, p. 351), many teachers and students criticised the
use of concordances due to an unfavourable attitude towards approach to working
with computers in general; for others, results displayed in the form of a classical concordance
are disturbing as it truncates sentences to centre the key word, which goes
against most students’ and teachers’ experience of promoting meaning analysis; corpora
represent an authentic language that may be beyond the scope of the learner’s
language knowledge; DDL activities require at least basic skills in corpus usage, and
are time-consuming because learners formulate their own rules based on data ob-
servation.
In my opinion, these misgivings may come from fear of the unknown, uncertainty
concerning the advantages of corpus methods in language learning, and nonuser-friendly
corpus tools (cf. Leńko-Szymańska, 2017). The benefits of concordance
displays lie in their focus on a particular language phenomenon; learners are not
distracted by having to read whole sentences, and they may restrict their observation
to the target phenomenon. Reading the whole sentence is a habit rather than
a necessity; we may often limit ourselves to the closest context to understand the
meaning. In addition, vertical reading helps to find more repeating patterns (such
as collocations) for a search word (cf. Boulton, 2009; Ballance, 2017). This approach
also forces learners to think about what they are learning, which has a positive influence
on retention of new knowledge in memory (Bernardini, 2004; Römer, 2011).
However, it must be admitted that authentic language in a corpus can be demotivating,
especially for beginners. Therefore, it is crucial to adjust corpus-based exercises
to target the learner group and provide ready-to-use activities that do not always
require direct work with a corpus and too much information. Thanks to that, each
user can easily choose whether to work with a computer or to do on-paper exercises.
Concordance is a key element here, as was already postulated in an early stage
of DDL (Johns, 1991, p. 2) and is still actively used in many studies (Bennett, 2010;Boulton,
2010; Liontou, 2020; Szudarski, 2020). However, nowadays, concordancebased
exercises are not the only solution; we may also work with frequency lists
OPEN
ACCESS
adrian jan zasina127
(cf. Nation, 2016), collocations (cf. Vyatkina, 2016), and many other functions provided
by corpus browsers and tools (cf. Zasina, 2018a).
The strong point of DDL lies in the authenticity of language usage and the inductive
approach that engages the learner in individual work and develops their independent
thinking. Moreover, this method makes it possible to adapt classroom materials
to the specific purposes for any group of learners. Therefore, the workbook
presented tries to deliver easy-to-use materials for teachers and students who can,
without much knowledge of corpus linguistics, simply start to use corpora in teaching
and learning.
3 DESIGN OF THE WORKBOOK
The workbook is suitable for university students of CFL. It was set based on exercises
that were prepared within an academic course on Grammatical and Lexical Corpusbased
Exercises at Charles University. The aim of the course was to teach students how
to use language data in language learning — DDL. The same procedure was adopted
in the workbook, where specific exercises guide students through three steps (identification,
classification, generalisation).
The workbook consists of hands-on and hands-off exercises (Boulton, 2012).
Hands-off exercises are suitable in a classroom without access to a computer. This
makes it possible to use corpus-based exercises in facilities without access to a computer
study room, thereby making the corpus approach more accessible. Hands-on
exercises cover direct corpus activities with corpus tools and classical gap filling exercises
that use corpus data displayed in the form of frequency lists or collocations.
There are also supplementary non-corpus exercises. Each exercise is marked by
a symbol indicating the type of exercise (direct corpus, indirect corpus, non-corpus).
Exercises that involve direct computer work are linked with a specific online tool that
is required to solve the exercises.
The workbook is designed to help teachers and students to choose topics that are
of interest. There is no need to proceed chapter by chapter. It is up to the user to plan
their own study programme. The workbook consists of a short introduction and more
than 100 exercises with a key. It is a collection of corpus-based exercises; therefore,
there are no detailed instructions on how to work with corpus data and corpus tools.
However, the introductory part features a short description of corpus tools with links
to manuals which are available on-line. There are two reasons to not include manuals
in the corpus workbook. First, they are already described and easily accessible in
Czech and English on the website https://wiki.korpus.cz/doku.php/manualy. Second,
corpus tools are under continuous development to improve their functionalities, and
it is better to get acquainted with the most current version. In addition, brand-new
tools can appear as well. Thus, each user of the workbook should first become familiar
with basic work with corpus tools. Fortunately, the direct exercises in the workbook
use user-friendly tools that are mostly intuitive.
Due to the authenticity of the corpus material, the workbook is recommended for
higher intermediate students starting from level B1, although some exercises could
OPEN
ACCESS
128STUDIE Z APLIKOVANÉ LINGVISTIKY 2/2022
be introduced to beginners as well; in that case, the teacher’s assistance is required.
Individual teachers know their students best and may select suitable corpus exercises
for their classes. It is also possible to modify a task to adjust its form to the students’
needs. Therefore, all exercises offered should be taken as a source for inspiration.
3.1 THEMATIC AREAS
The workbook is divided into 14 chapters. Each chapter represents a specific linguistic
issue focused on the following thematic areas: difference in meaning (word meaning
based on context), grammatical gender, declension, nominalised adjectives, hard and
soft adjectives, pronouns, prepositions, prefixes (s- vs. z-), past participle, collocability,
idioms, stylistic variants, the competing endings -a and -u in the genitive singular,
and vowel length. The topics were organised around the content of the academic
course Grammatical and Lexical Corpus-based Exercises, previous in-depth analysis of
the most problematic areas of foreign Czech in learner corpus (Zasina, 2019), experiences
with university teaching abroad (Zasina, 2022), and on my personal experience
with teachers’ needs.1
3.2 EXERCISE EXAMPLES
This subsection zooms in on exercise examples included in the corpus workbook. The
following tasks present two corpus-based exercises concerning the dative case, and
collocability. The exercises are on-computer activities (hands-on). They make use of
data from the SYN2020 corpus (Křen et al., 2020).
Exercise 1
This exercise concerning the dative case represents the chapter on declension. The
dative case is (along with the vocative) one of the less frequent cases in Czech; its
most common uses are to express the addressee as indirect object or a single object
governed by a verb requiring a dative construction (Cvrček et al., 2010, p. 139). There
are also several typical prepositions used in this task that require this case. Students
work with a provided list of the five most frequent dative prepositions as identified in
a corpus of contemporary written Czech. Their task is to search in the corpus for the
prepositions and find two example sentences for each of them. Through observing
the concordances, students have the opportunity to familiarise themselves with each
preposition and its meaning. Then, students share their ideas in pairs. In the final step,
the teacher discusses examples with the whole group to systematise the new material.
Task. Podívejte se na pět nejčastějších dativních předložek v níže uvedené tabulce. Zkuste je
vyhledat v korpusu současné češtiny a vypsat dvě příkladové věty pro každou z nich.
‘Take a look at the five most common dative prepositions in the table below. Try to
find them in a corpus of contemporary Czech and write two example sentences for
each of them.’
1	 Special thanks to Barbora Hrabalová for her comments and recommendations.
OPEN
ACCESS
adrian jan zasina129
No.
Předložka
‘Preposition’
Frekvence
‘Frequency’
1 k ‘to’ 545,155
2 proti ‘against’ 53,821
3 kvůli ‘due to’ 42,560
4 díky ‘thanks to’ 40,362
5 vůči ‘against’ 12,564
Table 1: Seznam nejčastějších dativních předložek na základě korpusu SYN2020 ‘List of the most frequent
dative prepositions based on the SYN2020 corpus’
Exercise 2
This exercise presents an example from a chapter dealing with collocability. All exercises
work with the co-occurrence of two words, mainly adjectives with corresponding
nouns. In this chapter, students can work with the Word at a Glance tool
(Machálek, 2020b), a word profile aggregator displaying collocations and many other
details. Alternatively, in the exercise below, students can search for collocations using
simple frequency lists or the collocation candidates function of the KonText corpus
browser (Machálek, 2020a). The following task exemplifies the use of the collocation
function. First, the students search the corpus for the nouns given in the exercise.
Second, they choose the “collocations > custom” option from the list in the KonText
browser to identify the most likely adjectival collocates. They can specify the parameters
of the collocation search; students must ensure that lemma is set as an attribute,
the collocation windows span is <–5, 5>, and results are sorted using the logDice collocation
measure. Thanks to this setup, they will be able to identify the most usual
collocations. The task helps students to understand the collocational behaviour of
the word and teaches them autonomy in learning. Teacher assistance is not required;
therefore, this exercise may be done as homework.
Task. Na základě dat z korpusu s použitím funkce KOLOKACE zjistěte, která níže uvedená
adjektiva se pojí s jednotlivými substantivy:
‘Using the COLLOCATIONS function, please use corpus data to identify which adjectives
co-occur with individual nouns:’
historický, žhavý, starý, širý, nutný, lidský, modrý, odborný, přirozený, dopravní
………….. nebe, ………….. podmínka, ………….. škola, ………….. jádro, ………….. prostředí,
………….. muž, ………….. nehoda, ………….. oči, ………….. novinka, ………….. život
4 CONCLUSION
Integrating corpus methods into regular teaching is a challenge, as it is a new approach
that requires a change in thinking, not only for teachers, but also for ­learners.
OPEN
ACCESS
130STUDIE Z APLIKOVANÉ LINGVISTIKY 2/2022
Teachers and learners need to know exactly what kind of advantages the DDL method
brings. Several studies (Vališová, 2012, 2016; Zasina, 2018b, 2019) have already explained
the role of DDL in CFL learning and proposed concrete corpus-based activities.
However, it is still necessary to promote this approach and to meet the requirements
and expectations of potential users. Furthermore, the possibility that
computers might not be available during regular classes should also be taken into account;
on-paper exercises could then be a useful introduction to DDL.
As Boulton and Cobb’s (2017, p. 386) meta-analysis concludes, “DDL works pretty
well in almost any context where it has been extensively tried”, therefore a corpus
workbook with a broad range of materials is an opportunity to take a first step into
the use of corpus in teaching and learning Czech. The workbook delivers ready-touse
comprehensive study materials for CFL students and their teachers. Although
previous publications (Šmejkalová & Kopřivová, 2019; Šormová et al., 2019) tried to
motivate teachers to use corpus methods in language teaching, they only focused on
teaching native speakers of Czech, and they did not address issues in CFL teaching.
This publication pays attention to the most prominent obstacles in the Czech language
for foreigners, which are different in many respects from the mistakes made by
native users. The rich offer of corpus-based exercises can provide a basis for university
course content dealing with the DDL method. It can also serve as an inspiration
for teachers to prepare their own corpus-based exercises meeting the requirements
of a specific group of learners. Last but not least, all corpus exercises are primarily
intended for CFL students to help them resolve difficulties and understand the principles
of the Czech language. This material is a starting point that needs continuing
development; therefore, I do believe it will provide an impulse that contributes to
popularising the DDL method in CFL teaching and learning.
Acknowledgements
This paper was supported by the project International mobility of research, technical and administrative
staff at the Charles University no CZ.02.2.69/0.0/0.0/18_053/0016976. I would also like to extend
my gratitude to Neil Bermel for helpful comments and proofreading.
REFERENCES
Ballance, O. J. (2017). Pedagogical models of
concordance use: Correlations between
concordance user preferences. Computer
Assisted Language Learning, 30(3–4), 259–283.
Bennett, G. R. (2010). Using Corpora in the
Language Learning Classroom: Corpus
Linguistics for Teachers. University of
Michigan Press.
Bernardini, S. (2004). Corpora in the classroom:
An overview and some reflections on future
developments. In J. McH. Sinclair (Ed.), How to
Use Corpora in Language Teaching (pp. 15–36).
John Benjamins.
Boulton, A. (2009). Testing the limits of datadriven
learning: Language proficiency and
training. ReCALL, 21(1), 37–54.
Boulton, A. (2010). Data-driven learning: Taking
the computer out of the equation. Language
Learning, 60(3), 534–572. DOI
Boulton, A. (2012). Hands-on / hands-off:
Alternative approaches to data-driven
learning. In J. Thomas & A. Boulton (Eds.),
OPEN
ACCESS
adrian jan zasina131
Input, Process and Product: Developments in
Teaching and Language Corpora (pp. 152–168).
Masaryk University Press.
Boulton, A., & Cobb, T. (2017). Corpus use in
language learning: A meta-analysis. Language
Learning, 67(2), 348–393. DOI
Čermák, F., Blatná, R., Klímová, J., Kopřivová,
M., Kučera, K., Petkevič, V., Schmiedtová, V.,
& Šulc, M. (2005). Jak využívat Český národní
korpus. Nakladatelství Lidové noviny.
Cvrček, V., Kodýtek, V., Kopřivová, M.,
Kováříková, D., Sgall, P., Šulc, M., Táborský, J.,
Volín, J., & Waclawičová, M. (2010). Mluvnice
současné češtiny 1: Jak se píše a jak se mluví.
Karolinum.
Johns, T. (1991). Should you be persuaded: Two
samples of data-driven learning materials.
Classroom Concordancing: ELR Journal, 4, 1–16.
Kočařová, B. (2013). Korpus a jmenný rod ve
výuce češtiny jako cizího jazyka [Bachelor’s
thesis]. Masaryk Universisty, Faculty of Arts.
Available at https://is.muni.cz/th/fvi7f/.
Konečná, H., & Zasina, A. J. (2014). Studium
českého jazyka a internet. In E. Rusinová (Ed.),
Přednášky a besedy ze XLVII. běhu LŠSS (pp.
104–112). Masaryk University.
Křen, M., Cvrček, V., Henyš, J., Hnátková, M.,
Jelínek, T., Kocek, J., Kováříková, D., Křivan,
J., Milička, J., Petkevič, V., Procházka, P.,
Skoumalová, H., Šindlerová, J., & Škrabal,
M. (2020). SYN2020: Representative corpus of
written Czech. Institute of the Czech National
Corpus, Faculty of Arts, Charles University.
Available at www.korpus.cz.
Leńko-Szymańska, A. (2017). Training teachers
in data driven learning: Tackling the
challenge. Language Learning & Technology,
21(3), 217–241.
Liontou, T. (2020). The effect of datadriven
learning activities on young EFL
learners’ processing of English idioms. In
P. Crosthwaite (Ed.), Data-Driven Learning for
the Next Generation: Corpora and DDL for Pretertiary
Learners (pp. 208–225). Routledge.
Lukšija, M. (2010). Korpus jako zdroj dat při
prezentaci předložek do/na s místním směrovým
významem ve výuce češtiny pro cizince
[Bachelor’s thesis]. Masaryk University,
Faculty of Arts. Available at https://is.muni.
cz/th/217240/ff_b/.
Lukšija, M. (2012). Korpusy a česká deklinace ve
výuce češtiny jako cizího jazyka [Master’s thesis].
Masaryk University, Faculty of Arts. Available
at https://is.muni.cz/th/217240/ff_m/.
Machálek, T. (2020a). Kontext: Advanced and
flexible corpus query interface. In N. Calzolari
et al. (Eds.), Proceedings of The 12th Language
Resources and Evaluation Conference
(pp. 7003–7008). European Language
Resources Association.
Machálek, T. (2020b). Word at a Glance:
Modular word profile aggregator. In
N. Calzolari et al. (Eds.), Proceedings of The
12th Language Resources and Evaluation
Conference (pp. 7009–7014). European
Language Resources Association.
Nation, I. S. P. (2016). Making and Using Word
Lists for Language Learning and Testing. John
Benjamins Publishing Company.
Osolsobě, K. (2010). Jak se učit česky
s korpusem. In E. Rusinová (Ed.), Přednášky
a besedy z XLIII. běhu LŠSS (pp. 112–119).
Masaryk University.
Römer, U. (2011). Corpus research applications
in second language teaching. Annual Review of
Applied Linguistics, 31, 205–225. DOI
Šmejkalová, M., & Kopřivová, M. (2019).
Jazykové korpusy a elektronické zdroje ve výuce
českého jazyka. Charles University, Faculty of
Education. Available at https://futurebooks.
cz/books/pedfa_esf_5/?/obalka/.
Šormová, K., Šebesta, K., Gráf, T., Hejhalová, V.,
& Kukrechtová, B. (2019). Korpusy v jazykovém
vyučování. Charles University, Faculty of Arts.
Šulc, M. (2007). Experimentujeme s češtinou. Jak
pracovat s korpusem českého jazyka ve školách
i mimo ně. Nakladatelství Lidové noviny.
Szudarski, P. (2020). Effects of data-driven
learning on enhancing the phraseological
knowledge of secondary school learners of L2
English. In P. Crosthwaite (Ed.), Data-Driven
Learning for the Next Generation: Corpora and
DDL for Pre-tertiary Learners (pp. 133–149).
Routledge.
Tribble, C. (2015). Teaching and language
corpora: Perspectives from a personal journey.
OPEN
ACCESS
132STUDIE Z APLIKOVANÉ LINGVISTIKY 2/2022
In A. Leńko-Szymańska & A. Boulton (Eds.),
Multiple Affordances of Language Corpora
for Data-driven Learning (pp. 37–62). John
Benjamins.
Vališová, P. (2009). Korpus jako zdroj dat
systémového popisu české konjugace při výuce
češtiny jako cizího jazyka [Master’s thesis].
Masaryk University, Faculty of Arts. Available
at https://is.muni.cz/th/v98z0/.
Vališová, P. (2012). Data-driven learning a výuka
češtiny jako cizího jazyka. CASALC Review, 2,
22–39.
Vališová, P. (2016). Korpus ve výuce češtiny
jako cizího jazyka — typy cvičení. In I. Starý
Kořánová & T. Vučka (Eds.), Čeština jako cizí
jazyk VIII (pp. 128–141). Charles University.
Vyatkina, N. (2016). Data-driven learning
of collocations: Learner performance,
proficiency, and perceptions. Language
Learning & Technology, 20(3), 159–179. DOI
Zasina, A. J. (2018a). Korpusy językowe
w nauczaniu języków obcych — Metoda,
narzędzia, praktyka. In M. Jedynak (Ed.),
Specyficzne potrzeby studentów szkół wyższych
a nauczanie języków obcych. Tom II. Praktyczne
narzędzia (pp. 110–123). Studium Praktycznej
Nauki Języków Obcych Uniwersytetu
Wrocławskiego.
Zasina, A. J. (2018b). O problémech nerodilých
mluvčích s kvantitou na základě analýzy
korpusových dat. In S. Škodová & M. Hrdlička
(Eds.), Čeština jako cizí jazyk v průsečíku pohledů
(pp. 281–298). Charles University, Faculty of
Arts.
Zasina, A. J. (2019). Korpusový přístup
ve výuce češtiny jako cizího jazyka.
[Dissertation]. Charles University, Faculty
of Arts. Available at https://dspace.cuni.cz/
handle/20.500.11956/115540.
Zasina, A. J. (2022). Corpus approach to teaching
Czech as a foreign language in university
courses. Bohemistyka, XXII(3), 435–462. DOI
Adrian Jan Zasina | Institute of the Czech National Corpus,
Faculty of Arts, Charles University
<adrian.zasina@ff.cuni.cz>
School of Languages and Cultures, University of Sheffield
<a.zasina@sheffield.ac.uk>
OPEN
ACCESS