Presentations at DiGS 15 and PLC38 about priming effects in the history of English, by Aaron Ecay and Meredith Tamminga
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
data
figures
finished
queries
scripts
static-figures
.gitignore
LICENSE
README.md
biblio.bib
digs-slides.cls
digs15-presentation-handout.tex
digs15-presentation.tex
digs15-proceedings-extended-abstract.tex
plc38-presentation-handout.tex
plc38-presentation.tex

README.md

Introduction

This repository contains the source code for the presentation entitled “Persistence as a diagnostic of grammatical status” at DiGS15 by Aaron Ecay and Meredith Tamminga.

Instructions

Corpus data

Middle English negation

In order to replicate the corpus portion of this analysis, you will need a copy of the Penn Parsed Corpus of Middle English (PPCME2) and version 2.003.04 of the CorpusSearch program.

To preform the replication, you should examine the make-corpus.sh script in the queries subdirectory of this repository.

  1. Edit the line of the file beginning CS_COMMAND= so that it points to the location on your computer where the CorpusSearch program is.

  2. You will also need to place a file named ppcme2.out in the queries directory that contains the PPCME2 corpus (concatenating together the .psd files from the corpus release suffices).

Once you have performed these two steps, you should run the make-corpus.sh script. The final output of this process is the coding.cod.ooo file, a copy of which is also included in this repository for the convenience of those who do not have access to the PPCME2. This file is the input to the next stage.

Do-support

The do-support code relies on the data file do.dat in the data subdirectory of this repository. (Please contact the authors if you are interested in the scripts which generate this file from the PPCEME and PCEEC). The code for creating the priming graph is contained in the file dosupp.R in the scripts subdirectory.

Analysis

The analysis is provided in the form of R source code,, in two files in the scripts subdirectory. In order to use this code, you will need to install the stringr, ggplot2, plyr, reshape2, binom, and tikzDevice packages (the last only if you intend to replicate the graphs for compilation to PDF, and not merely interactively in the R console):

install.packages(c("stringr","plyr","ggplot2", "reshape2"))
install.packages(c("tikzDevice"), repos = c("http://R-Forge.R-project.org"))

(The tikzDevice package has to be installed from a non-default repository, because it is not distributed on CRAN, the main R package distribution network.)

You should set R’s working directory to the root of this repository (e.g. with the setwd() function.)

Once you have done that, load the two scripts:

source("scripts/data.R")
source("scripts/graphs.R")

Then, load the data into R:

neg <- cleanNegData()

Then, you can make the graphs:

all.graphs()

Inspect the source in the graphs.R file for more detail about individual graphs.

Slide show

In order to recreate the slide show from the LaTeX source, you will need several LaTeX packages; consult the presentation.tex and digs-slides.cls files for exact details. You should compile the slides using the lualatex program, and biber for the bibliography.

The easiest way to install all the necessary programs is the TeXlive distribution.

Comments

If you have comments on the presentation, the analysis, or any aspect of the work, please feel free to email ecay@ling.upenn.edu.