R notebook for the corpus-based study of English suffixes -ance and -ancy

Authors
Affiliations

Ni Ketut Budiani

Computer-assisted Lexicology and Lexicography (CompLexico) research group, CIRHSS, Udayana University

Bachelor of English Literature (BoEL), Faculty of Humanities, Udayana University

Gede Primahadi Wijaya Rajeg

Computer-assisted Lexicology and Lexicography (CompLexico) research group, CIRHSS, Udayana University

Bachelor of English Literature (BoEL), Faculty of Humanities, Udayana University

I Made Netra

Bachelor of English Literature (BoEL), Faculty of Humanities, Udayana University

Published

March 18, 2025

── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
✔ dplyr     1.1.4     ✔ readr     2.1.5
✔ forcats   1.0.0     ✔ stringr   1.5.1
✔ ggplot2   3.5.1     ✔ tibble    3.2.1
✔ lubridate 1.9.4     ✔ tidyr     1.3.1
✔ purrr     1.0.4     
── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
✖ dplyr::filter() masks stats::filter()
✖ dplyr::lag()    masks stats::lag()
ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errors

Attaching package: 'googledrive'


The following objects are masked from 'package:googlesheets4':

    request_generate, request_make

Overview

As of the 28th of May 2025, the article, titled “Productivity and Distribution of Nominal Suffixes -ance and -ancy in the Corpus of Contemporary American English” is under review in LET: Linguistics, Literature and English Teaching Journal.

Data preparation

# df <- read_sheet(ss = datasheet) <- run this regularly to check update
# write_rds(df, "data-raw/ance-ancy.rds")
df <- read_rds("data-raw/ance-ancy.rds")
unanalysed <- read_tsv("data-raw/unanalysed.csv") |> 
  mutate(SUFFIX = replace_na(SUFFIX, ""))
Rows: 279 Columns: 5
── Column specification ────────────────────────────────────────────────────────
Delimiter: "\t"
chr (5): FORM, ROOT, SUFFIX, ETYMOLOGY, REMARK

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.
unanalysed1 <- unanalysed |> 
  filter(!is.na(ROOT), SUFFIX != "-mancy") |> 
  filter(FORM != "CIRCUMSTANCE") |> # historical relation with circum + stance
  mutate(SUFFIX = replace(SUFFIX, SUFFIX == "", "ancy")) |> 
  select(FORM, ROOT_new = ROOT, SUFFIX_new = SUFFIX, ETYMOLOGY_new = ETYMOLOGY)
# df_combined <- df |> 
#   left_join(unanalysed1,
#             by = join_by(FORM))|> 
#   mutate(ROOT = if_else(!is.na(ROOT_new),
#                         ROOT_new,
#                         ROOT),
#          ETYMOLOGY = if_else(!is.na(ETYMOLOGY_new),
#                              ETYMOLOGY_new,
#                              ETYMOLOGY),
#          SUFFIX = if_else(!is.na(SUFFIX_new),
#                           SUFFIX_new,
#                           SUFFIX)) |> 
#   select(!matches("_new"))

df_combined <- df

Try the current analysed data

df_ok <- filter(df_combined, !is.na(ROOT))
# df_ok <- filter(df, !is.na(ROOT))
df_ok |> slice_sample(n = 10)
# A tibble: 10 × 7
      ID FORM         FREQ GENRE     SUFFIX ROOT        ETYMOLOGY               
   <dbl> <chr>       <dbl> <chr>     <chr>  <chr>       <chr>                   
 1    28 MENDICANCY     44 ACADEMIC  ancy   mendicant   mendicancy is formed wi…
 2     3 PERFORMANCE 12767 BLOG      ance   perform     performance is formed w…
 3    86 BIBLIOMANCY     1 NEWSPAPER ancy   bibliomancy bibliomancy is formed w…
 4    50 RELEVANCE     456 NEWSPAPER ance   relevant    relevance is formed wit…
 5    15 INERRANCY     123 WEBSITE   ancy   inerrant    inerrancy is formed wit…
 6    58 RESEMBLANCE   222 SPOKEN    ance   resemble    resemblance is a borrow…
 7    77 SEMBLANCE     240 MAGAZINE  ance   semblance   semblance is a borrowin…
 8    58 NUISANCE      358 NEWSPAPER ance   nuisant     nuisance is a borrowing…
 9    36 INCONSTANCY     3 TV/MOVIES ancy   inconstant  inconstancy is a borrow…
10    61 DEFIANCE      416 MAGAZINE  ance   defiant     defiance is a borrowing…

Productivity analysis

Run productivity analysis per suffix (Table 1).

# drive_create("productivity-by-affix",
#              path = as_id("1e_0dwXEH_qTTzUdDGNxI_LHQSFIS98qj"),
#              type = "spreadsheet")
# Created Drive file:
# • productivity-by-affix <id: 1g9fw5PVUrSiR9TGA9tQlOBPektaZJaVMqBK7_wwAIOs>
# With MIME type:
# • application/vnd.google-apps.spreadsheet


prod_by_affix <- df_ok |> 
  
  # remove GENRE to re-calculate frequency of suffixes and their hapax
  select(-GENRE) |> 
  
  # to sum token frequency
  group_by(FORM, SUFFIX, ROOT) |> 
  summarise(n_token = sum(FREQ), .groups = "drop") |> 
  
  # determine the hapax
  mutate(is_hapax = if_else(n_token == 1, TRUE, FALSE)) |> 
  
  # to run productivity analysis
  group_by(SUFFIX) |> 
  summarise(n_type = n_distinct(FORM),
            n_token = sum(n_token),
            n_hapax = sum(is_hapax),
            hapax_per_token_ratio = n_hapax/n_token)

prod_by_affix |> 
  knitr::kable(col.names = c("Suffix", "Type Freq.", "Token Freq.", "No. Hapax", "Hapax per Token Ratio"))
# prod_by_affix |> 
#   rename(Suffix = SUFFIX,
#          `Type Freq.` = n_type,
#          `Token Freq.` = n_token,
#          `No. of Hapax` = n_hapax,
#          `Hapax per token ratio` = hapax_per_token_ratio) |> 
#   write_sheet(ss = "1g9fw5PVUrSiR9TGA9tQlOBPektaZJaVMqBK7_wwAIOs",
#               sheet = "Sheet1")

write_tsv(prod_by_affix, file = "data-out/productivity_overall_by_affix.tsv")
Table 1: Productivity measures of -ance and -ancy
Suffix Type Freq. Token Freq. No. Hapax Hapax per Token Ratio
ance 97 823125 0 0.0000000
ancy 103 48601 11 0.0002263

Run productivity analysis by genres using the type frequency measure (Figure 1).

ance <- df_ok |> 
  filter(SUFFIX == "ance")

ance_word_freq <- ance |> 
  group_by(FORM) |> 
  summarise(FREQ = sum(FREQ))

ancy <- df_ok |> 
  filter(SUFFIX == "ancy")

ancy_word_freq <- ancy |> 
  group_by(FORM) |> 
  summarise(FREQ = sum(FREQ))

ance_prod <- ance |> 
  group_by(GENRE) |> 
  
  # determine the hapax in a given genre
  mutate(is_hapax = if_else(FREQ == 1, TRUE, FALSE)) |> 
  summarise(n_type = n_distinct(FORM),
            n_token = sum(FREQ),
            n_hapax = sum(is_hapax)) |> 
  ungroup() |> 
  mutate(suffix = "ancy")

ancy_prod <- ancy |> 
  group_by(GENRE) |> 
  
  # determine the hapax in a given genre
  mutate(is_hapax = if_else(FREQ == 1, TRUE, FALSE)) |> 
  summarise(n_type = n_distinct(FORM),
            n_token = sum(FREQ),
            n_hapax = sum(is_hapax)) |> 
  ungroup() |> 
  mutate(suffix = "ance")

genre_prod <- bind_rows(ance_prod, ancy_prod)
genre_prod |> 
  mutate(suffix = str_c("-", suffix, sep = "")) |> 
  ggplot(aes(x = GENRE, y = n_type, fill = suffix)) + 
  geom_col(position = position_dodge(width = .9)) +
  coord_flip() +
  theme_light(base_family = "serif") +
  labs(y = "Type frequency",
       fill = "Suffix",
       x = "Genre") +
  theme(legend.text = element_text(size = 13),
        legend.title = element_text(size = 17),
        axis.title.y = element_text(size = 17),
        axis.title.x = element_text(size = 17),
        axis.text.x = element_text(size = 13)) +
  scale_fill_discrete(breaks = c("-ancy", "-ance"))
# ggsave("figures/prod-by-genre.png", width = 6.5, height = 4.5, dpi = 300,
#        units = "in")

# googledrive::drive_upload("figures/prod-by-genre.png",
#                           path = as_id("18ppX_-Nr-dI-iQojYC5WHpGZi_dKojhA"),
#                           name = "productivity-by-genre.png")
Figure 1: Type Frequency of each suffix across the COCA’s genres

Interchangeability of the base/root

Check the shared and distinct bases (Table 2). This is operationalised via stripping off the -ance and -ancy strings from the FORM column.

ance_form <- ance_word_freq |> 
  mutate(BASE = str_replace(FORM, "ANCE$", ""),
         suffix = "ance")

ancy_form <- ancy_word_freq |> 
  mutate(BASE = str_replace(FORM, "ANCY$", ""),
         suffix = "ancy")

all_ance_ancy <- bind_rows(ance_form, ancy_form)

all_bases <- unique(c(ance_form$BASE, ancy_form$BASE))

shared_ance_ancy <- intersect(ance_form$BASE, ancy_form$BASE)
shared_ance_ancy_prop <- round((length(shared_ance_ancy)/length(all_bases)) * 100, 2)

only_ance <- setdiff(ance_form$BASE, ancy_form$BASE)
only_ance_prop <- round((length(only_ance)/length(all_bases)) * 100, 2)

only_ancy <- setdiff(ancy_form$BASE, ance_form$BASE)
only_ancy_prop <- round((length(only_ancy)/length(all_bases)) * 100, 2)

Out of the total 183 bases, only 9.29% (i.e., 17 bases) are shared (i.e., appear with -ance and -ancy) but the frequency of occurrence of these shared bases with the suffixes are not equal.

all_ance_ancy_tb <- all_ance_ancy |> 
  filter(BASE %in% shared_ance_ancy) |> 
  select(-FORM) |> 
  pivot_wider(names_from = "suffix", values_from = "FREQ") |> 
  mutate(BASE = str_c(BASE, "ANC(E/Y)", sep = "")) |> 
  arrange(desc(ancy))

write_tsv(all_ance_ancy_tb, "data-out/all_ance_ancy_tb.tsv")

all_ance_ancy_tb |> 
  arrange(BASE) |> 
  knitr::kable()
Table 2: Token frequencies of the shared base between each suffix
BASE ance ancy
ARROGANC(E/Y) 5049 5
BRILLIANC(E/Y) 3203 86
COMPLIANC(E/Y) 12987 9
CONSTANC(E/Y) 1974 594
CONTINUANC(E/Y) 805 6
DEVIANC(E/Y) 323 112
EXTRAVAGANC(E/Y) 458 2
FRAGRANC(E/Y) 3055 3
IGNORANC(E/Y) 13310 1
IRRELEVANC(E/Y) 624 171
PREDOMINANC(E/Y) 417 2
RADIANC(E/Y) 1075 2
RELEVANC(E/Y) 8094 589
RELUCTANC(E/Y) 4560 3
RESISTANC(E/Y) 32578 1
RESONANC(E/Y) 3919 3
SIGNIFICANC(E/Y) 21828 6

Binomial test

all_ance_ancy_tb_binom <- all_ance_ancy_tb |> 
  mutate(binom = map2(ance, 
                      ancy, 
                      ~binom.test(.x,
                                  sum(c(.x, .y)),
                                  .5))) |> 
  mutate(binom = map(binom, ~tidy(.))) |> 
  unnest_wider(binom) |> 
  mutate(p_holm = p.adjust(p.value, method = "holm")) |> 
  arrange(BASE) |> 
  mutate(across(matches("(^estimate|^conf)"), ~round(., digits = 4)))
all_ance_ancy_tb_binom |> 
  knitr::kable()
write_tsv(all_ance_ancy_tb_binom, file = "data-out/interchangeable-binom-test.tsv")
Table 3: Binomial tests for the shared base distribution between -ance and -ancy
BASE ance ancy estimate statistic p.value parameter conf.low conf.high method alternative p_holm
ARROGANC(E/Y) 5049 5 0.9990 5049 0 5054 0.9977 0.9997 Exact binomial test two.sided 0
BRILLIANC(E/Y) 3203 86 0.9739 3203 0 3289 0.9678 0.9790 Exact binomial test two.sided 0
COMPLIANC(E/Y) 12987 9 0.9993 12987 0 12996 0.9987 0.9997 Exact binomial test two.sided 0
CONSTANC(E/Y) 1974 594 0.7687 1974 0 2568 0.7519 0.7849 Exact binomial test two.sided 0
CONTINUANC(E/Y) 805 6 0.9926 805 0 811 0.9840 0.9973 Exact binomial test two.sided 0
DEVIANC(E/Y) 323 112 0.7425 323 0 435 0.6987 0.7830 Exact binomial test two.sided 0
EXTRAVAGANC(E/Y) 458 2 0.9957 458 0 460 0.9844 0.9995 Exact binomial test two.sided 0
FRAGRANC(E/Y) 3055 3 0.9990 3055 0 3058 0.9971 0.9998 Exact binomial test two.sided 0
IGNORANC(E/Y) 13310 1 0.9999 13310 0 13311 0.9996 1.0000 Exact binomial test two.sided 0
IRRELEVANC(E/Y) 624 171 0.7849 624 0 795 0.7547 0.8130 Exact binomial test two.sided 0
PREDOMINANC(E/Y) 417 2 0.9952 417 0 419 0.9829 0.9994 Exact binomial test two.sided 0
RADIANC(E/Y) 1075 2 0.9981 1075 0 1077 0.9933 0.9998 Exact binomial test two.sided 0
RELEVANC(E/Y) 8094 589 0.9322 8094 0 8683 0.9267 0.9374 Exact binomial test two.sided 0
RELUCTANC(E/Y) 4560 3 0.9993 4560 0 4563 0.9981 0.9999 Exact binomial test two.sided 0
RESISTANC(E/Y) 32578 1 1.0000 32578 0 32579 0.9998 1.0000 Exact binomial test two.sided 0
RESONANC(E/Y) 3919 3 0.9992 3919 0 3922 0.9978 0.9998 Exact binomial test two.sided 0
SIGNIFICANC(E/Y) 21828 6 0.9997 21828 0 21834 0.9994 0.9999 Exact binomial test two.sided 0

Reuse

Citation

BibTeX citation:
@software{budiani2025,
  author = {Budiani, Ni Ketut and Rajeg, Gede Primahadi Wijaya and
    Netra, I Made},
  title = {R Notebook for the Corpus-Based Study of {English} Suffixes
    -\emph{Ance} and -\emph{Ancy}},
  date = {2025-03-18},
  url = {https://complexico.github.io/ance-ancy-suffix/},
  doi = {10.17605/OSF.IO/7YJ49},
  langid = {en}
}
For attribution, please cite this work as:
Budiani, Ni Ketut, Gede Primahadi Wijaya Rajeg, and I Made Netra. 2025. “R Notebook for the Corpus-Based Study of English Suffixes -Ance and -Ancy.” https://doi.org/10.17605/OSF.IO/7YJ49.