Methods of Corpus Linguistics

Introduction to the course

Mariana Montes

Outline

• Topics
• Software to install
• Format of the classes
• Exam and submission procedure

Topics

Corpus linguistics

• What is a corpus?

• What is corpus linguistics?

• Collocations and keywords analysis

• Comparison of variants and of varieties

Statistical techniques

• Association measures

Comparison of variants

• Logistic regression

• Conditional inference trees

Comparison of varieties

• Correspondence Analysis

• Factor Analysis

Software to install

Programs

• R – the programming language

• RStudio – IDE for R

• Quarto – publishing system

• Tinytex (I will show you how)

• (Optionally) Git

Costs?

Everything open source and free!

R packages from CRAN

After installing R, you should install the following R packages.

• tidyverse (a group of data manipulation packages)
• ca (for Correspondence Analysis)
• here (for file paths)
• xml2 (to work with XML files)
• easystats (for reporting stats)¹
• ggeffects (to plot regression effects)

1. You might also want to run easystats::install_suggested() to install additional useful packages

R development packages

• mclm (“masterclm/mclm”)
• mclmtutorials (“masterclm/mclmtutorials”)
• learnr (“rstudio/learnr”)
• gradethis (“rstudio/gradethis”)
• (Optionally) glossr (“montesmariana/glossr”) if you want to write interlinear glosses

How to install R packages

• From CRAN: with install.packages("package")

install.packages(c("tidyverse", "ca", "here", "xml2", "easystats", "ggeffects"))

• Development packages: with remotes::install_github("user/repo")

library(remotes)
install_github("masterclm/mclm")
install_github("masterclm/mclmtutorials")
install_github("rstudio/learnr")
install_github("rstudio/gradethis")

Run this code in the R console.

Format of the classes

Technical setup

• Basics of R, RStudio, R projects

• Basics of Quarto (to write your paper!)

• Basics of Git (option to submit the paper)

Theoretical classes

Lectures on the different topics, with slides

• Corpus linguistics

• Association measures

• Logistic regression and conditional trees

• Correspondence Analysis

• Factor Analysis

Case studies

Going through analyses, with notebooks and code you can copy-paste

• Collocation and keyword analysis

• Alternation studies: analysis of variants

• Lectometry: analysis of varieties

• Register analysis: analysis of varieties

Exam and submission procedure

Exam format

• Paper with analysis

  • Choose a corpus and at least one technique

  • Define a research question that the technique can address

  • Small literature review

  • Perform analysis

• Full project to be submitted: R code, paper written in Quarto, bibliography

Submission procedure

Option 1: Toledo

• Turn your project folder into a zip file (excluding corpus) and submit via Toledo Assignments

Option 2: Git & GitHub

• Follow the instructions to set up the Git repository and push your project

• Optional intermediate assignments to get used to the tasks, Git and GitHub

Git submission - setup

• Create an R project with a git repository

• Stage and commit your changes, either in the Git tab or by typing git add . followed by git commit -m "some message" in the Git Bash Terminal

• Add a remote to the repository by typing git remote add origin <url> in the Git Bash Terminal

• Set the main branch to “main” with git branch -M main if it isn’t already.

• Upload your changes with git push -u origin main

Git submission - later

• Stage your changes with git add . (. to stage everything)

• Take a snapshot of your work with git commit -m "some message"

• When you want to submit, run git push

Basics of R