Introduction

This tutorial introduces how to extract concordances and keyword-in-context (KWIC) displays with R.

This tutorial is aimed at beginners and intermediate users of R with the aim of showcasing how to extract keywords and key phrases from textual data and how to process the resulting concordances using R. The aim is not to provide a fully-fledged analysis but rather to show and exemplify selected useful methods associated with concordancing.

The entire R Notebook for the tutorial can be downloaded here. If you want to render the R Notebook on your machine, i.e. knitting the document to html or a pdf, you need to make sure that you have R and RStudio installed and you also need to download the bibliography file and store it in the same folder where you store the Rmd file.

Binder
Click this link to open an interactive version of this tutorial on MyBinder.org.
This interactive Jupyter notebook allows you to execute code yourself and you can also change and edit the notebook, e.g. you can change code and upload your own data.


In the language sciences, concordancing refers to the extraction of words from a given text or texts (Lindquist 2009, 5). Commonly, concordances are displayed in the form of keyword-in-context displays (KWICs) where the search term is shown in context, i.e. with preceding and following words. Concordancing are central to analyses of text and they often represents the first step in more sophisticated analyses of language data (Stefanowitsch 2020). The play such a key role in the language sciences because concordances are extremely valuable for understanding how a word or phrase is used, how often it is used, and in which contexts is used. As concordances allow us to analyze the context in which a word or phrase occurs and provide frequency information about word use, they also enable us to analyze collocations or the collocational profiles of words and phrases (Stefanowitsch 2020, 50–51). Finally, concordances can also be used to extract examples and it is a very common procedure.

\label{fig:Fig1} Concordances in AntConc.

Concordances in AntConc.


There are various very good software packages that can be used to create concordances - both for offline use (e.g. AntConc (Anthony 2004), SketchEngine(Kilgarriff et al. 2004), MONOCONC(Barlow 1999), and ParaConc)(Barlow 2002) and online use (see e.g. here).

In addition, many corpora that are available such as the BYU corpora can be accessed via a web interface that have in-built concordancing functions.

\label{fig:Fig2} Online concordances extracted from the COCA corpus that is part of the BYU corpora.

Online concordances extracted from the COCA corpus that is part of the BYU corpora.

While these packages are very user-friendly, offer various additional functionalities, and almost everyone who is engaged in analyzing language has used concordance software, they all suffer from shortcomings that render R a viable alternative. Such issues include that these applications

  • are black boxes that researchers do not have full control over or do not know what is going on within the software

  • they are not open source

  • they hinder replication because the replications is more time consuming compared to analyses based on Notebooks.

  • they are commonly not free-of charge or have other restrictions on use (a notable exception is AntConc)

R represents an alternative to ready-made concordancing applications because it:

  • is extremely flexible and enables researchers to perform their entire analysis in a single environment

  • allows full transparency and documentation as analyses can be based on Notebooks

  • offer version control measures (this means that the specific versions of the involved software are traceable)

  • makes research more replicable as entire analyses can be reproduced by simply running the Notebooks that the research is based on

Especially the aspect that R enables full transparency and replicability is relevant given the ongoing Replication Crisis (Yong 2018; Aschwanden 2018; Diener and Biswas-Diener 2019; Velasco 2019; McRae 2018). The Replication Crisis is a ongoing methodological crisis primarily affecting parts of the social and life sciences beginning in the early 2010s (see also Fanelli 2009). Replication is important so that other researchers, or the public for that matter, can see or, indeed, reproduce, exactly what you have done. Fortunately, R allows you to document your entire workflow as you can store everything you do in what is called a script or a notebook (in fact, this document was originally a R notebook). If someone is then interested in how you conducted your analysis, you can simply share this notebook or the script you have written with that person.

Preparation and session set up

This tutorial is based on R. If you have not installed R or are new to it, you will find an introduction to and more information how to use R here. For this tutorials, we need to install certain packages from an R library so that the scripts shown below are executed without errors. Before turning to the code below, please install the packages by running the code below this paragraph. If you have already installed the packages mentioned below, then you can skip ahead and ignore this section. To install the necessary packages, simply run the following code - it may take some time (between 1 and 5 minutes to install all of the libraries so you do not need to worry if it takes some time).

# install packages
install.packages("quanteda")
install.packages("dplyr")
install.packages("stringr")
install.packages("flextable")
# install klippy for copy-to-clipboard button in code chunks
install.packages("remotes")
remotes::install_github("rlesur/klippy")

Now that we have installed the packages, we activate them as shown below.

# activate packages
library(quanteda)
library(dplyr)
library(stringr)
library(flextable)
# activate klippy for copy-to-clipboard button
klippy::klippy()

Once you have installed R and RStudio and also initiated the session by executing the code shown above, you are good to go.

Loading and processing textual data

For this tutorial, we will use Lewis Caroll’s Alice’s Adventures in Wonderland. You can use the code below to load this text into R (but you have to have access to the internet to do so).

text <- base::readRDS(url("https://slcladal.github.io/data/alice.rda", "rb"))

The table above shows that the example text requires formatting so that we can use it. Therefore, we collapse it into a single object (or text) and remove superfluous white spaces.

text <- text %>%
  # collapse lines into a single  text
  paste0(collapse = " ") %>%
  # remove superfluous white spaces
  str_squish()

The result confirms that the entire text is now combined into a single character object.

Creating simple concordances

Now that we have loaded the data, we can easily extract concordances using the kwic function from the quanteda package. The kwic function takes the text (x) and the search pattern (pattern) as it main arguments but it also allows the specification of the context window, i.e. how many words/elements are show to the left and right of the key word (we will go over this later on).

kwic_alice <- kwic(
  # define text
  text, 
  # define search pattern
  pattern = "Alice")

You will see that you get a warning stating that you should use token before extracting concordances. This can be done as shown below. Also, we can specify the package from which we want to use a function by adding the package name plus :: before the function (see below)

kwic_alice <- quanteda::kwic(
  # define and tokenize text
  quanteda::tokens(text), 
  # define search pattern
  pattern = "alice")

We can easily extract the frequency of the search term (alice) using the nrow or the length functions which provide the number of rows of a tables (nrow) or the length of a vector (length).

nrow(kwic_alice)
## [1] 386
length(kwic_alice$keyword)
## [1] 386

The results show that there are 414 instances of the search term (alice) but we can also find out how often different variants (lower case versus upper case) of the search term were found using the table function. This is especially useful when searches involve many different search terms (while it is, admittedly, less useful in the present example).

table(kwic_alice$keyword)
## 
## Alice 
##   386

To get a better understanding of the use of a word, it is often useful to extract more context. This is easily done by increasing size of the context window. To do this, we specify the window argument of the kwic function. In the example below, we set the context window size to 10 words/elements rather than using the default (which is 5 word/elements).

kwic_alice_longer <- kwic(
  # define text
  text, 
  # define search pattern
  pattern = "alice", 
  # define context window size
  window = 10)

EXERCISE TIME!

`

  1. Extract the first 10 concordances for the word confused.
Answer
  kwic_confused <- kwic(x = text, pattern = "confused")
  ## Warning: 'kwic.character()' is deprecated. Use 'tokens()' first.
  # inspect
  kwic_alice %>%
  as.data.frame() %>%
  head(10)
  ##    docname from  to                         pre keyword
  ## 1    text1   14  14    I . Down the Rabbit-Hole   Alice
  ## 2    text1   73  73          a book , " thought   Alice
  ## 3    text1  153 153           in that ; nor did   Alice
  ## 4    text1  239 239       and then hurried on ,   Alice
  ## 5    text1  309 309 In another moment down went   Alice
  ## 6    text1  348 348     down , so suddenly that   Alice
  ## 7    text1  531 531          " Well ! " thought   Alice
  ## 8    text1  657 657             for , you see ,   Alice
  ## 9    text1  731 731                got to ? " (   Alice
  ## 10   text1  924 924             else to do , so   Alice
  ##                                   post pattern
  ## 1            was beginning to get very   alice
  ## 2  " without pictures or conversations   alice
  ## 3              think it so _very_ much   alice
  ## 4                started to her feet ,   alice
  ## 5                after it , never once   alice
  ## 6                  had not a moment to   alice
  ## 7                 to herself , " after   alice
  ## 8         had learnt several things of   alice
  ## 9            had no idea what Latitude   alice
  ## 10          soon began talking again .   alice
  1. How many instances are there of the word wondering?
Answer
  kwic(x = text, pattern = "wondering") %>%
  as.data.frame() %>%
  nrow()
  ## Warning: 'kwic.character()' is deprecated. Use 'tokens()' first.
  ## [1] 7
  1. Extract concordances for the word strange and show the first 5 concordance lines.
Answer
  kwic_strange <- kwic(x = text, pattern = "strange")
  ## Warning: 'kwic.character()' is deprecated. Use 'tokens()' first.
  # inspect
  kwic_strange %>%
  as.data.frame() %>%
  head(5)
  ##   docname  from    to                          pre keyword
  ## 1   text1  3547  3547 her voice sounded hoarse and strange
  ## 2   text1 13273 13273         , that it felt quite strange
  ## 3   text1 33247 33247    remember them , all these strange
  ## 4   text1 33458 33458    her became alive with the strange
  ## 5   text1 33784 33784        and eager with many a strange
  ##                               post pattern
  ## 1              , and the words did strange
  ## 2               at first ; but she strange
  ## 3      Adventures of hers that you strange
  ## 4 creatures of her little sister's strange
  ## 5         tale , perhaps even with strange

`


Extracting more than single words

While extracting single words is very common, you may want to extract more than just one word. To extract phrases, all you need to so is to specify that the pattern you are looking for is a phrase, as shown below.

kwic_pooralice <- kwic(text, pattern = phrase("poor alice"))

You may also want to extract more or less fixed patterns rather than exact words or phrases. To search for patterns that allow variation rather than specific, exactly-defined words, you need to include regular expressions in your search pattern.


EXERCISE TIME!

`

  1. Extract the first 10 concordances for the phrase the hatter.
Answer
  kwic_thehatter <- kwic(x = text, pattern = phrase("the hatter"))
  ## Warning: 'kwic.character()' is deprecated. Use 'tokens()' first.
  # inspect
  kwic_thehatter %>%
  as.data.frame() %>%
  head(10)
  ##    docname  from    to                    pre    keyword
  ## 1    text1 16710 16711   wish I'd gone to see the Hatter
  ## 2    text1 16741 16742 and the March Hare and the Hatter
  ## 3    text1 16993 16994 wants cutting , " said the Hatter
  ## 4    text1 17039 17040     it's very rude . " The Hatter
  ## 5    text1 17187 17188         a bit ! " said the Hatter
  ## 6    text1 17312 17313      with you , " said the Hatter
  ## 7    text1 17347 17348  , which wasn't much . The Hatter
  ## 8    text1 17425 17426  days wrong ! " sighed the Hatter
  ## 9    text1 17476 17477         in as well , " the Hatter
  ## 10   text1 17591 17592 should it ? " muttered the Hatter
  ##                           post    pattern
  ## 1      instead ! " CHAPTER VII the hatter
  ## 2        were having tea at it the hatter
  ## 3        . He had been looking the hatter
  ## 4    opened his eyes very wide the hatter
  ## 5           . " You might just the hatter
  ## 6  , and here the conversation the hatter
  ## 7       was the first to break the hatter
  ## 8               . " I told you the hatter
  ## 9   grumbled : " you shouldn't the hatter
  ## 10       . " Does _your_ watch the hatter
  1. How many instances are there of the phrase the hatter?
Answer
  kwic_thehatter %>%
  as.data.frame() %>%
  nrow()
  ## [1] 51
  1. Extract concordances for the phrase the cat and show the first 5 concordance lines.
Answer
  kwic_thecat <- kwic(x = text, pattern = phrase("the cat"))
  ## Warning: 'kwic.character()' is deprecated. Use 'tokens()' first.
  # inspect
  kwic_thecat %>%
  as.data.frame() %>%
  head(5)
  ##   docname  from    to               pre keyword                     post
  ## 1   text1   946   947   ! " ( Dinah was the cat             . ) " I hope
  ## 2   text1 15756 15757 a few yards off . The Cat only grinned when it saw
  ## 3   text1 15881 15882   get to , " said the Cat         . " I don't much
  ## 4   text1 15907 15908   you go , " said the Cat            . " - so long
  ## 5   text1 15937 15938  do that , " said the Cat          , " if you only
  ##   pattern
  ## 1 the cat
  ## 2 the cat
  ## 3 the cat
  ## 4 the cat
  ## 5 the cat

`


Searches using regular expressions

Regular expressions allow you to search for abstract patterns rather than concrete words or phrases which provides you with an extreme flexibility in what you can retrieve. A regular expression (in short also called regex or regexp) is a special sequence of characters that stand for are that describe a pattern. You can think of regular expressions as very powerful combinations of wildcards or as wildcards on steroids. For example, the sequence [a-z]{1,3} is a regular expression that stands for one up to three lower case characters and if you searched for this regular expression, you would get, for instance, is, a, an, of, the, my, our, etc, and many other short words as results.

There are three basic types of regular expressions:

  • regular expressions that stand for individual symbols and determine frequencies

  • regular expressions that stand for classes of symbols

  • regular expressions that stand for structural properties

The regular expressions below show the first type of regular expressions, i.e. regular expressions that stand for individual symbols and determine frequencies.

The regular expressions below show the second type of regular expressions, i.e. regular expressions that stand for classes of symbols.

The regular expressions that denote classes of symbols are enclosed in [] and :. The last type of regular expressions, i.e. regular expressions that stand for structural properties are shown below.

To include regular expressions in your KWIC searches, you include them in your search pattern and set the argument valuetype to "regex". The search pattern "\\balic.*|\\bhatt.*" retrieves elements that contain alic and hatt followed by any characters and where the a in alic and the h in hatt are at a word boundary, i.e. where they are the first letters of a word. Hence, our search would not retrieve words like malice or shatter. The | is an operator (like +, -, or *) that stands for or.

# define search patterns
patterns <- c("\\balic.*|\\bhatt.*")
kwic_regex <- kwic(
  # define text
  text, 
  # define search pattern
  patterns, 
  # define valuetype
  valuetype = "regex")

EXERCISE TIME!

`

  1. Extract the first 10 concordances for words containing exu.
Answer
  kwic_exu <- kwic(x = text, pattern = ".*exu.*", valuetype = "regex")
  ## Warning: 'kwic.character()' is deprecated. Use 'tokens()' first.
  # inspect
  kwic_exu %>%
  as.data.frame() %>%
  head(10)
  ## [1] docname from    to      pre     keyword post    pattern
  ## <0 Zeilen> (oder row.names mit Länge 0)
  1. How many instances are there of words beginning with pit?
Answer
  kwic(x = text, pattern = "\\bpit.*", valuetype = "regex") %>%
  as.data.frame() %>%
  nrow()
  ## Warning: 'kwic.character()' is deprecated. Use 'tokens()' first.
  ## [1] 5
  1. Extract concordances for words ending with ption and show the first 5 concordance lines.
Answer
  kwic(x = text, pattern = "ption\\b", valuetype = "regex")  %>%
  as.data.frame() %>%
  head(5)
  ## Warning: 'kwic.character()' is deprecated. Use 'tokens()' first.
  ##   docname from   to                         pre  keyword
  ## 1   text1 5823 5823 adjourn , for the immediate adoption
  ##                           post  pattern
  ## 1 of more energetic remedies - ption\\b

`


Piping concordances

Quite often, we only want to retrieve patterns if they occur in a certain context. For instance, we might be interested in instances of selection but only if the preceding word is natural. Such conditional concordances could be extracted using regular expressions but they are easier to retrieve by piping. Piping is done using the %>% function from the dplyr package and the piping sequence can be translated as and then. We can then filter those concordances that contain natural using the filter function from the dplyr package. Note the the $ stands for the end of a string so that natural$ means that natural is the last element in the string that is preceding the keyword.

kwic_pipe <- kwic(x = text, pattern = "alice") %>%
  dplyr::filter(stringr::str_detect(pre, "poor$|little$"))

Piping is a very useful helper function and it is very frequently used in R - not only in the context of text processing but in all data science related domains.

Arranging concordances and adding frequency information

When inspecting concordances, it is useful to re-order the concordances so that they do not appear in the order that they appeared in the text or texts but by the context. To reorder concordances, we can use the arrange function from the dplyr package which takes the column according to which we want to re-arrange the data as it main argument.

In the example below, we extract all instances of alice and then arrange the instances according to the content of the post column in alphabetical.

kwic_ordered <- kwic(x = text, pattern = "alice") %>%
  dplyr::arrange(post)

Arranging concordances according to alphabetical properties may, however, not be the most useful option. A more useful option may be to arrange concordances according to the frequency of co-occurring terms or collocates. In order to do this, we need to extract the co-occurring words and calculate their frequency. We can do this by combining the mutate, group_by, n() functions from the dplyr package with the str_remove_all function from the stringr package. Then, we arrange the concordances by the frequency of the collocates in descending order (that is why we put a - in the arrange function). In order to do this, we need to

  1. create a new variable or column which represents the word that co-occurs with, or, as in the example below, immediately follows the search term. In the example below, we use the mutate function to create a new column called post_word. We then use the str_remove_all function to remove everything except for the word that immediately follows the search term (we simply remove everything and including a white space).

  2. group the data by the word that immediately follows the search term.

  3. create a new column called post_word_freq which represents the frequencies of all the words that immediately follow the search term.

  4. arrange the concordances by the frequency of the collocates in descending order.

kwic_ordered_coll <- kwic(
  # define text
  x = text, 
  # define search pattern
  pattern = "alice") %>%
  # extract word following the keyword
  dplyr::mutate(post_word = str_remove_all(post, " .*")) %>%
  # group following words
  dplyr::group_by(post_word) %>%
  # extract frequencies of the following words
  dplyr::mutate(post_word_freq = n()) %>%
  # arrange/order by the frequency of the following word
  dplyr::arrange(-post_word_freq)

We add more columns according to which we could arrange the concordance following the same schema. For example, we could add another column that represented the frequency of words that immediately preceded the search term and then arrange according to this column.

Ordering by subsequent elements

In this section, we will extract the three words following the keyword (alice) and organize the concordances by the frequencies of the following words. We begin by inspecting the first 6 lines of the concordance of selection.

head(kwic_alice)
## Keyword-in-context with 6 matches.                                                   
##   [text1, 14]     I. Down the Rabbit-Hole | Alice |
##   [text1, 73]            a book," thought | Alice |
##  [text1, 153]            in that; nor did | Alice |
##  [text1, 239]        and then hurried on, | Alice |
##  [text1, 309] In another moment down went | Alice |
##  [text1, 348]      down, so suddenly that | Alice |
##                                     
##  was beginning to get very          
##  " without pictures or conversations
##  think it so _very_ much            
##  started to her feet,               
##  after it, never once               
##  had not a moment to

Next, we take the concordances and create a clean post column that is all in lower case and that does not contain any punctuation.

kwic_alice %>%
  # convert to data frame
  as.data.frame() %>%
  # create new CleanPost
  dplyr::mutate(CleanPost = stringr::str_remove_all(post, "[:punct:]"),
                CleanPost = stringr::str_squish(CleanPost),
                CleanPost = tolower(CleanPost))-> kwic_alice_following
# inspect
head(kwic_alice_following)
##   docname from  to                         pre keyword
## 1   text1   14  14    I . Down the Rabbit-Hole   Alice
## 2   text1   73  73          a book , " thought   Alice
## 3   text1  153 153           in that ; nor did   Alice
## 4   text1  239 239       and then hurried on ,   Alice
## 5   text1  309 309 In another moment down went   Alice
## 6   text1  348 348     down , so suddenly that   Alice
##                                  post pattern                         CleanPost
## 1           was beginning to get very   alice         was beginning to get very
## 2 " without pictures or conversations   alice without pictures or conversations
## 3             think it so _very_ much   alice             think it so very much
## 4               started to her feet ,   alice               started to her feet
## 5               after it , never once   alice               after it never once
## 6                 had not a moment to   alice               had not a moment to

In a next step, we extract the 1st, 2nd, and 3rd words following the keyword.

kwic_alice_following %>%
  # extract first element after keyword
  dplyr::mutate(FirstWord = stringr::str_remove_all(CleanPost, " .*")) %>%
  # extract second element after keyword
  dplyr::mutate(SecWord = stringr::str_remove(CleanPost, ".*? "),
                SecWord = stringr::str_remove_all(SecWord, " .*")) %>%
  # extract third element after keyword
  dplyr::mutate(ThirdWord = stringr::str_remove(CleanPost, ".*? "),
                ThirdWord = stringr::str_remove(ThirdWord, ".*? "),
                ThirdWord = stringr::str_remove_all(ThirdWord, " .*")) -> kwic_alice_following
# inspect
head(kwic_alice_following)
##   docname from  to                         pre keyword
## 1   text1   14  14    I . Down the Rabbit-Hole   Alice
## 2   text1   73  73          a book , " thought   Alice
## 3   text1  153 153           in that ; nor did   Alice
## 4   text1  239 239       and then hurried on ,   Alice
## 5   text1  309 309 In another moment down went   Alice
## 6   text1  348 348     down , so suddenly that   Alice
##                                  post pattern                         CleanPost
## 1           was beginning to get very   alice         was beginning to get very
## 2 " without pictures or conversations   alice without pictures or conversations
## 3             think it so _very_ much   alice             think it so very much
## 4               started to her feet ,   alice               started to her feet
## 5               after it , never once   alice               after it never once
## 6                 had not a moment to   alice               had not a moment to
##   FirstWord   SecWord ThirdWord
## 1       was beginning        to
## 2   without  pictures        or
## 3     think        it        so
## 4   started        to       her
## 5     after        it     never
## 6       had       not         a

Next, we calculate the frequencies of the subsequent words and order in descending order from the 1st to the 3rd word following the keyword.

kwic_alice_following %>%
  # calculate frequency of following words
  # 1st word
  dplyr::group_by(FirstWord) %>%
  dplyr::mutate(FreqW1 = n()) %>%
  # 2nd word
  dplyr::group_by(SecWord) %>%
  dplyr::mutate(FreqW2 = n()) %>%
  # 3rd word
  dplyr::group_by(ThirdWord) %>%
  dplyr::mutate(FreqW3 = n()) %>%
  # ungroup
  dplyr::ungroup() %>%
  # arrange by following words
  dplyr::arrange(-FreqW1, -FreqW2, -FreqW3) -> kwic_alice_following
# inspect results
head(kwic_alice_following, 10)
## # A tibble: 10 × 14
##    docname  from    to pre         keyword post  pattern Clean…¹ First…² SecWord
##    <chr>   <int> <int> <chr>       <chr>   <chr> <fct>   <chr>   <chr>   <chr>  
##  1 text1   15840 15840 "so far , … Alice   ", a… alice   and sh… and     she    
##  2 text1   20942 20942 "be behead… Alice   ", a… alice   and sh… and     she    
##  3 text1   25847 25847 "quite a n… Alice   ", a… alice   and sh… and     she    
##  4 text1   33229 33229 "curious d… Alice   ", a… alice   and sh… and     she    
##  5 text1   33350 33350 ", and thi… Alice   "and… alice   and al… and     all    
##  6 text1   16498 16498 "said pig … Alice   "; \… alice   and i … and     i      
##  7 text1    3625  3625 "words , \… Alice   ", a… alice   and he… and     her    
##  8 text1    1692  1692 "here befo… Alice   ", )… alice   and ro… and     round  
##  9 text1   25955 25955 "eyes . He… Alice   ", a… alice   and tr… and     tried  
## 10 text1    6573  6573 "you know … Alice   ", \… alice   and wh… and     why    
## # … with 4 more variables: ThirdWord <chr>, FreqW1 <int>, FreqW2 <int>,
## #   FreqW3 <int>, and abbreviated variable names ¹​CleanPost, ²​FirstWord

The results now show the concordance arranged by the frequency of the words following the keyword.

Concordances from transcriptions

As many analyses use transcripts as their primary data and because transcripts have features that require additional processing, we will now perform concordancing based on on transcripts. As a first step, we load five example transcripts that represent the first five files from the Irish component of the International Corpus of English.

# define corpus files
files <- paste("https://slcladal.github.io/data/ICEIrelandSample/S1A-00", 1:5, ".txt", sep = "")
# load corpus files
transcripts <- sapply(files, function(x){
  x <- readLines(x)
  })

The first ten lines shown above let us know that, after the header (<S1A-001 Riding>) and the symbol which indicates the start of the transcript (<I>), each utterance is preceded by a sequence which indicates the section, file, and speaker (e.g. <S1A-001$A>). The first utterance is thus uttered by speaker A in file 001 of section S1A. In addition, there are several sequences that provide meta-linguistic information which indicate the beginning of a speech unit (<#>), pauses (<,>), and laughter (<&> laughter </&>).

To perform the concordancing, we need to change the format of the transcripts because the kwic function only works on character, corpus, tokens object- in their present form, the transcripts represent a list which contains vectors of strings. To change the format, we collapse the individual utterances into a single character vector for each transcript.

transcripts_collapsed <- sapply(files, function(x){
  # read-in text
  x <- readLines(x)
  # paste all lines together
  x <- paste0(x, collapse = " ")
  # remove superfluous white spaces
  x <- str_squish(x)
})

We can now extract the concordances.

kwic_trans <- quanteda::kwic(
  # tokenize transcripts
  quanteda::tokens(transcripts_collapsed), 
  # define search pattern
  pattern = phrase("you know"))

The results show that each non-alphanumeric character is counted as a single word which reduces the context of the keyword substantially. Also, the docname column contains the full path to the data which make it hard to parse the content of the table. To address the first issue, we specify the tokenizer that we will use to not disrupt the annotation too much. In addition, we clean the docname column and extract only the file name. Lastly, we will expand the context window to 10 so that we have a better understanding of the context in which the phrase was used.

kwic_trans <- quanteda::kwic(
  # tokenize transcripts
  quanteda::tokens(transcripts_collapsed, what = "fasterword"), 
  # define search
  pattern = phrase("you know"),
  # extend context
  window = 10) %>%
  # clean docnames
  dplyr::mutate(docname = str_replace_all(docname, ".*/([A-Z][0-9][A-Z]-[0-9]{1,3}).txt", "\\1"))

Extending the context can also be used to identify the speaker that has uttered the search pattern that we are interested in. We will do just that as this is a common task in linguistics analyses.

To extract speakers, we need to follow these steps:

  1. Create normal concordances of the pattern that we are interested in.

  2. Generate concordances of the pattern that we are interested in with a substantially enlarged context window size.

  3. Extract the speakers from the enlarged context window size.

  4. Add the speakers to the normal concordances using the left-join function from the dplyr package.

kwic_normal <- quanteda::kwic(
  # tokenize transcripts
  quanteda::tokens(transcripts_collapsed, what = "fasterword"), 
  # define search
  pattern = phrase("you know")) %>%
  as.data.frame()
kwic_speaker <- quanteda::kwic(
    # tokenize transcripts
  quanteda::tokens(transcripts_collapsed, what = "fasterword"), 
  # define search
  pattern = phrase("you know"), 
  # extend search window
  window = 500) %>%
  # convert to data frame
  as.data.frame() %>%
  # extract speaker (comes after $ and before >)
  dplyr::mutate(speaker = stringr::str_replace_all(pre, ".*\\$(.*?)>.*", "\\1")) %>%
  # extract speaker
  dplyr::pull(speaker)
# add speaker to normal kwic
kwic_combined <- kwic_normal %>%
  # add speaker
  dplyr::mutate(speaker = kwic_speaker) %>%
  # simplify docname
  dplyr::mutate(docname = stringr::str_replace_all(docname, ".*/([A-Z][0-9][A-Z]-[0-9]{1,3}).txt", "\\1")) %>%
  # remove superfluous columns
  dplyr::select(-to, -from, -pattern)

The resulting table shows that we have successfully extracted the speakers (identified by the letters in the speaker column) and cleaned the file names (in the docnames column).

Customizing concordances

As R represents a fully-fledged programming environment, we can, of course, also write our own, customized concordance function. The code below shows how you could go about doing so. Note, however, that this function only works if you enter more than a single file.

mykwic <- function(txts, pattern, context) {
  # activate packages
  require(stringr)
  # list files
  txts <- txts[stringr::str_detect(txts, pattern)]
  conc <- sapply(txts, function(x) {
    # determine length of text
        lngth <- as.vector(unlist(nchar(x)))
    # determine position of hits
    idx <- str_locate_all(x, pattern)
    idx <- idx[[1]]
    ifelse(nrow(idx) >= 1, idx <- idx, return(NA))
    # define start position of hit
    token.start <- idx[,1]
    # define end position of hit
    token.end <- idx[,2]
    # define start position of preceding context
    pre.start <- ifelse(token.start-context < 1, 1, token.start-context)
    # define end position of preceding context
    pre.end <- token.start-1
    # define start position of subsequent context
    post.start <- token.end+1
    # define end position of subsequent context
    post.end <- ifelse(token.end+context > lngth, lngth, token.end+context)
    # extract the texts defined by the positions
    PreceedingContext <- substring(x, pre.start, pre.end)
    Token <- substring(x, token.start, token.end)
    SubsequentContext <- substring(x, post.start, post.end)
    Id <- 1:length(Token)
    conc <- cbind(Id, PreceedingContext, Token, SubsequentContext)
    # return concordance
    return(conc)
    })
  concdf <- do.call(rbind, conc) %>%
    as.data.frame()
  return(concdf)
}

We can now try if this function works by searching for the sequence you know in the transcripts that we have loaded earlier. One difference between the kwic function provided by the quanteda package and the customized concordance function used here is that the kwic function uses the number of words to define the context window, while the mykwic function uses the number of characters or symbols instead (which is why we use a notably higher number to define the context window).

kwic_youknow <- mykwic(transcripts_collapsed, "you know", 50)

As this concordance function only works for more than one text, we split the text into chapters and assign each section a name.

# read in text
text_split <- text %>%
  stringr::str_squish() %>%
  stringr::str_split("[CHAPTER]{7,7} [XVI]{1,7}\\. ") %>%
  unlist()
text_split <- text_split[which(nchar(text_split) > 2000)]
# add names
names(text_split) <- paste0("text", 1:length(text_split))
# inspect data
nchar(text_split)
##  text1  text2  text3  text4  text5  text6  text7  text8  text9 text10 text11 
##  11331  10888   9137  13830  11767  13730  12564  13585  12527  11287  10292 
## text12 
##  11518

Now that we have named elements, we can search for the pattern poor alice. We also need to clean the concordance as some sections do not contain any instances of the search pattern. To clean the data, we select only the columns File, PreceedingContext, Token, and SubsequentContext and then remove all rows where information is missing.

mykwic_pooralice <- mykwic(text_split, "poor Alice", 50)

You can go ahead and modify the customized concordance function to suit your needs.

Citation & Session Info

Schweinberger, Martin. 2022. Concordancing with R. Brisbane: The University of Queensland. url: https://ladal.edu.au/kwics.html (Version 2022.11.15).

@manual{schweinberger2022kwics,
  author = {Schweinberger, Martin},
  title = {Concordancing with R},
  note = {https://ladal.edu.au/kwics.html},
  year = {2022},
  organization = "The University of Queensland, Australia. School of Languages and Cultures},
  address = {Brisbane},
  edition = {2011.11.15}
}
sessionInfo()
## R version 4.2.1 RC (2022-06-17 r82510 ucrt)
## Platform: x86_64-w64-mingw32/x64 (64-bit)
## Running under: Windows 10 x64 (build 19043)
## 
## Matrix products: default
## 
## locale:
## [1] LC_COLLATE=German_Germany.utf8  LC_CTYPE=German_Germany.utf8   
## [3] LC_MONETARY=German_Germany.utf8 LC_NUMERIC=C                   
## [5] LC_TIME=German_Germany.utf8    
## 
## attached base packages:
## [1] stats     graphics  grDevices utils     datasets  methods   base     
## 
## other attached packages:
## [1] flextable_0.8.2 stringr_1.4.1   dplyr_1.0.10    quanteda_3.2.2 
## 
## loaded via a namespace (and not attached):
##  [1] tidyselect_1.1.2   xfun_0.32          bslib_0.4.0        purrr_0.3.4       
##  [5] lattice_0.20-45    vctrs_0.4.1        generics_0.1.3     htmltools_0.5.3   
##  [9] yaml_2.3.5         base64enc_0.1-3    utf8_1.2.2         rlang_1.0.4       
## [13] jquerylib_0.1.4    pillar_1.8.1       glue_1.6.2         DBI_1.1.3         
## [17] gdtools_0.2.4      uuid_1.1-0         lifecycle_1.0.1    zip_2.2.0         
## [21] evaluate_0.16      knitr_1.40         fastmap_1.1.0      fansi_1.0.3       
## [25] highr_0.9          Rcpp_1.0.9         cachem_1.0.6       RcppParallel_5.1.5
## [29] jsonlite_1.8.0     systemfonts_1.0.4  fastmatch_1.1-3    stopwords_2.3     
## [33] digest_0.6.29      stringi_1.7.8      grid_4.2.1         cli_3.3.0         
## [37] tools_4.2.1        magrittr_2.0.3     sass_0.4.2         klippy_0.0.0.9500 
## [41] tibble_3.1.8       pkgconfig_2.0.3    ellipsis_0.3.2     Matrix_1.5-1      
## [45] data.table_1.14.2  xml2_1.3.3         assertthat_0.2.1   rmarkdown_2.16    
## [49] officer_0.4.4      rstudioapi_0.14    R6_2.5.1           compiler_4.2.1

Back to top

Back to HOME


References

Anthony, Laurence. 2004. “AntConc: A Learner and Classroom Friendly, Multi-Platform Corpus Analysis Toolkit.” Proceedings of IWLeL, 7–13.
Aschwanden, Christie. 2018. “Psychology’s Replication Crisis Has Made the Field Better.” https://fivethirtyeight.com/features/psychologys-replication-crisis-has-made-the-field-better/.
Barlow, Michael. 1999. “Monoconc 1.5 and Paraconc.” International Journal of Corpus Linguistics 4 (1): 173–84.
———. 2002. “ParaConc: Concordance Software for Multilingual Parallel Corpora.” In Proceedings of the Third International Conference on Language Resources and Evaluation. Workshop on Language Resources in Translation Work and Research, 20–24.
Diener, Edward, and Robert Biswas-Diener. 2019. “The Replication Crisis in Psychology.” https://nobaproject.com/modules/the-replication-crisis-in-psychology.
Fanelli, Daniele. 2009. “How Many Scientists Fabricate and Falsify Research? A Systematic Review and Meta-Analysis of Survey Data.” PLoS One 4: e5738.
Kilgarriff, Adam, Pavel Rychly, Pavel Smrz, and David Tugwell. 2004. “Itri-04-08 the Sketch Engine.” Information Technology 105: 116.
Lindquist, Hans. 2009. Corpus Linguistics and the Description of English. Edinburgh: Edinburgh University Press.
McRae, Mike. 2018. “Science’s ’Replication Crisis’ Has Reached Even the Most Respectable Journals, Report Shows.” https://www.sciencealert.com/replication-results-reproducibility-crisis-science-nature-journals.
Stefanowitsch, Anatol. 2020. Corpus Linguistics. A Guide to the Methodology. Textbooks in Language Sciences. Berlin: Language Science Press.
Velasco, Emily. 2019. “Researcher Discusses the the Science Replication Crisis.” https://phys.org/news/2018-11-discusses-science-replication-crisis.html.
Yong, Ed. 2018. “Psychology’s Replication Crisis Is Running Out of Excuses. Another Big Project Has Found That Only Half of Studies Can Be Repeated. And This Time, the Usual Explanations Fall Flat.” https://www.theatlantic.com/science/archive/2018/11/psychologys-replication-crisis-real/576223/.