1 of 46

Corpus-based �Constructional Analysis��Session1

Names of the teachers and universities involved

2 of 46

Taking Europe to the next level

Je propose la création d’universités européennes qui seront un réseau d’universités de plusieurs pays d’Europe ...

Ich schlage die Schaffung von europäischen Universitäten vor, die ein Netzwerk von Universitäten aus mehreren europäischen Ländern sein werden ...

I propose the creation of European universities which will be a network of universities from several European countries ...

Emmanuel Macron 2017

  • Real unification of the European Higher Education Area
  • Not just "international"isation, but truly European joint learning and teaching

3 of 46

DIONE

Digitalising Mobility and International Networks with Open Education

  • HUBerlin, UCLouvain, Belgrade, Oslo, Granada, Wolverhampton, Nauci.Me
  • Development of a new format for teaching collaboration: Micro-collaboration
  • Focus on humanities (linguistics) and on digital research skills

4 of 46

Micro-collaboration

  • Short, flexible collaboration - administratively simple, can be implemented by anyone
  • Focus on student collaboration
  • Goal:

Empower all teachers

Democratise internationalisation

Make internationalisation massive and seamless

5 of 46

MICRO-COLLABORATION PRESENTATION� �

6 of 46

Common course project

Micro-collaboration between University 1 & University 2 on Corpus-based Constructional Analysis

Collaboration between lecturers and students!

Organizational challenges: xxx

4 online sessions

of about 2h

Dates of the sessions

Course language: multilingual; slides in xxx with presentations in xxx

7 of 46

Organization of the 4 sessions

Session 1

Introduction

Getting to know

Presentation of the case study

 Learning outcomes

Introduction to the corpus tool (SketchEngine)

    • How to formulate corpus queries
    • How to create and download a random corpus sample

Session 2

How to use Excel for data annotation and analysis

Selection of relevant criteria for the corpus analysis 

Session 3

Data annotation

Q&A session: discussion of methodological problems and difficulties in the corpus analysis

Session 4

Group presentations and discussion of the results

8 of 46

Organization of the 4 sessions

Session 1

Introduction

Getting to know

Presentation of the case study

(Learning outcomes)

Introduction to the corpus tool (SketchEngine)

    • How to formulate corpus queries
    • How to create and download a random corpus sample

Session 2

How to use Excel for data annotation and analysis

Selection of relevant criteria for the corpus analysis 

Session 3

Data annotation

Q&A session: discussion of methodological problems and difficulties in the corpus analysis

Session 4

Group presentations and discussion of the results

9 of 46

GETTING TO KNOW...

10 of 46

We are different...

Activity 1

    • in groups (random but not mixed)
    • Choose ONE word to describe your country and briefly explain your choice 

logo of University 1

logo of University 2

11 of 46

But we are also the same!

Activity 2

    • in groups (random and mixed)
    • List as many points as possible that are common to all members of your group 

logo of University 1

logo of University 2

12 of 46

Organization of the 4 sessions

Session 1

Introduction

Getting to know

Presentation of the case study

(Learning outcomes)

Introduction to the corpus tool (SketchEngine)

    • How to formulate corpus queries
    • How to create and download a random corpus sample

Session 2

How to use Excel for data annotation and analysis

Selection of relevant criteria for the corpus analysis 

Session 3

Data annotation

Q&A session: discussion of methodological problems and difficulties in the corpus analysis

Session 4

Group presentations and discussion of the results

13 of 46

Case study

Construction: FORM ↔ ‘meaning’ 

    • Dutch: [een schat van een kind] ↔ ‘a lovely child’ (lit. ‘a treasure of a child’)
    • German: [ein Bär von einem Mann] ↔ ‘a very strong man’ (lit. ‘a bear of a man’)
    • Croatian: [čudo od djeteta] ↔ ‘a  talented child’ (lit. ‘a wonder of a child’)

Example of a common construction between Dutch, German and Croatian:

14 of 46

Case study: corpus examples

  • DUTCH
    • Ik vind dit een pracht van een gedicht en neem het mee in mijn gedachten. �'I think this is a beauty of a poem and take it with me in my thoughts.'
    • Of u nu Aruba, Curaçao of Bonaire bezoekt, met een vliegticket Midden-Amerika en de Carribbean bent u verzekerd van een droom van een vakantie. �'Whether you're visiting Aruba, Curaçao or Bonaire, with a Central America and the Carribbean flight ticket, you're assured of a dream vacation.'

15 of 46

Case study: corpus examples

  • GERMAN
    • Tatsächlich ist der Dreck, wie er oft genannt wird, in Wirklichkeit ein Wunder von einem Mikrobiotop.�'In fact, the dirt, as it is often called, is actually a miracle of a microbiotope.'
    • Ferdinand war ein Riese von einem Kerl, mit Bärenkräften.'Ferdinand was a giant of a guy, with the strength of a bear.'

16 of 46

Case study: corpus examples

  • CROATIAN
    • Ne bojte (sic!) se ovih nula od ljudi što samo zlobno komentiraju. �'Don't be afraid of these nobodies who only comment insidiously.'
    • Pato je dolazio u Milan kao nekakav mesija, čudo od igrača kojem će se svijet godinama gledati i diviti.'Pato has come to Milan like some kind of messiah, a miracle of a player whom the world will watch and admire for years.'

17 of 46

TASK

Which languages do you speak?

Do you find other examples of this construction in any of the languages you speak?

18 of 46

Case study

  • Frequent lemmas in N1 position :

Dutch

German

Croatian

pracht ‘beauty’

Bild ‘picture’

čudo ‘wonder‘

dijk ‘dike’

Bär ‘bear’

duša ‘soul‘

schat ‘treasure’

Traum ‘dream’

govno ‘shit‘

joekel ‘whopper’

Berg ‘mountain’

smeće ‘rubbish‘

droom ‘dream’

Schatz ‘treasure’

nula ‘zero‘

19 of 46

Special properties of the pattern

Which elements are fixed and which are variable?

    • Determiner 1?
    • N1?
    • Preposition?
    • Determiner 2?
    • N2?

What is the (semantic) head noun of the pattern?

What is the function of the other noun?

Can the meaning of the pattern be derived from the meaning of its separate components?

20 of 46

INTRODUCTION INTO CONSTRUCTION GRAMMAR

Some introductory videos:

21 of 46

Why this case study?

Interesting from a cross-linguistic perspective:

    • The construction is productive in Dutch, German and Croatian, which allows for interlingual comparisons

1

Interesting from a constructionist perspective:

    •  You will have to find out the typical and atypical properties of the construction at the form and meaning levels

2

Interesting from a corpus-based perspective:

    • The corpus study will allow you to analyze and compare the formal and semantic properties of the construction, and its productivity

3

22 of 46

Organization of the 4 sessions

Session 1

Introduction

Getting to know

Presentation of the case study

Learning outcomes

Introduction to the corpus tool (SketchEngine)

    • How to formulate corpus queries
    • How to create and download a random corpus sample

Session 2

How to use Excel for data annotation and analysis

Selection of relevant criteria for the corpus analysis 

Session 3

Data annotation

Q&A session: discussion of methodological problems and difficulties in the corpus analysis

Session 4

Group presentations and discussion of the results

23 of 46

��LEARNING OUTCOMES AND COMPETENCES� �

24 of 46

Concrete outcomes 

After this course you will be able to...

    • conduct a corpus research in Sketch Engine
    • sort and annotate examples in an Excel-file
    • compare the findings between different languages 
    • report the results of a corpus study

25 of 46

Competences

After these 4 sessions, students will have developed competences in different areas:

  • Some general competences  about empirical research methods, e.g.:

  • Several competences of the Research cycle, e.g.:

 I can evaluate empirical research; I can explain the research cycle to others. 

To some extent: I understand the limits of research; I understand different perspectives in science 

26 of 46

Competences

27 of 46

Competences

I research a limited topic in at least two languages using digital tools. 

28 of 46

Competences

I can identify my information needs and, with guidance, access secondary and selected primary sources with support using an appropriate research strategy. 

I know a limited number of qualitative and quantitative methods. I can apply some methods independently. 

I understand that the implementation has to follow the design and the two are in correlation

I strictly follow my design in implementation and understand that implementation can be a cycle

29 of 46

Competences

I collect empirical data consistently with digital tools. I know how to work around bias and observer paradox and do so with guidance. 

I independently analyse my own simple data of selected data types according to given patterns, observing the research design.

I draw conclusions from my analysis within the framework of the given theory under guidance. 

I independently draw conclusions from my analysis within the framework of a self-selected theory.

30 of 46

Competences

I can explain simple facts correctly in terms of terminology and with further knowledge and defend my point of view.

I can explain complex issues, explain and defend my point of view on them and reflect on the theoretical premises of my argumentation.

31 of 46

Organization of the 4 sessions

Session 1

Introduction

Getting to know

Presentation of the case study

(Learning outcomes)

Introduction to the corpus tool (SketchEngine)

    • How to formulate corpus queries
    • How to create and download a random corpus sample

Session 2

How to use Excel for data annotation and analysis

Selection of relevant criteria for the corpus analysis 

Session 3

Data annotation

Q&A session: discussion of methodological problems and difficulties in the corpus analysis

Session 4

Group presentations and discussion of the results

FIND

PLAN

IMPLEMENT

SHARE

32 of 46

Organization of the 4 sessions

Session 1

Introduction

Getting to know

Presentation of the case study

(Learning outcomes)

Introduction to the corpus tool (SketchEngine)

    • How to formulate corpus queries
    • How to create and download a random corpus sample

Session 2

How to use Excel for data annotation and analysis

Selection of relevant criteria for the corpus analysis 

Session 3

Data annotation

Q&A session: discussion of methodological problems and difficulties in the corpus analysis

Session 4

Group presentations and discussion of the results

33 of 46

 INTRODUCTION TO THE CORPUS TOOL�

34 of 46

Corpus login

  • Sketch Engine: https://auth.sketchengine.eu
  • Institutional login 
    • Catholic University of Louvain 
    • Humboldt-Universität zu Berlin
  • Institutional username and password
  • Corpus: 
    • Dutch Web 2014 (nlTenTen14)
    • German Web 2018 (deTenTen18)
    • Croatian Web (hrWaC  2.2, ReLDI)
    • Albanian web 2022 (Albanian Corpus)

34

35 of 46

Sketch Engine: login

35

36 of 46

Select corpus

36

37 of 46

Simple query

37

1

2

3

38 of 46

CQL query 

38

1

2

3

39 of 46

Sample

39

40 of 46

Random sample

40

2

1

41 of 46

Download 

41

2

1

3

42 of 46

CORPUS TOOL

Video presentations on the formulation of queries in SketchEngine are available on Moodle

43 of 46

TASK

Dutch​

German​

Croatian​

+ pracht ('beauty')​

+ Bild ('picture')​

+čudo ('wonder') ​

+ dijk ('dike')​

+ Bär ('bear')​

+ duša ('soul') ​

+schat ('treasure')​

+ Traum ('dream')​

+ govno ('shit') ​

+joekel ('whopper')​

+ Berg ('mountain')​

+smeće ('rubbish') ​

+ droom ('dream')​

+ Hüne ('giant')​

+ nula ('zero') ​

+ kanjer ('whopper')​

+ Schrank('wardrobe')​

+gromada ('giant')​

+ draak ('dragon')​

+ Baum ('tree')​

….​

+ parel ('pearl')​

+Schatz ('treasure')​

+ knaller ('stunner')​

+ Kerl ('guy')​

+ wolk ('cloud')​

+ Koloss('colossus')​

...​

...​

Look at the following examples and try to find out how they could be ranged into different semantic categories 

44 of 46

Three semantic categories

Positive connotation ('good')​

Negative connotation ('bad')​

Augmentative function �('big')​

'wonder'​

'dragon'​

'giant'​

'soul'​

'monster'​

'​tree'

'dream'​

'shit'​

​'bear'

'beauty'​

​'rubbish'

​'wardrobe'

'treasure'

'zero'

'mountain'

45 of 46

TASK

  • Group assignment
    • 3 groups: 1 group per semantic category
    • Mixed groups UCLouvain-HUBerlin
  • Step 1 
    • Create a query to check if the 5 items of your semantic category are used in the N1 position of the Dutch/Croatian/German/Albanian cxn
      • HUBerlin: 2 languages (German + X)
      • UCLouvain: 1 language (Dutch), but syntactic + morphological cxn in Dutch 
  • Step 2 
    • Create and download a random sample of 500 occurrences per language/per construction 
    • Which items are used in which language? 

46 of 46

SEE YOU NEXT WEEK!