Data from: Decolonizing the ourang-outang
Data files
Dec 13, 2022 version files 207.04 KB
-
british_corpus_compilation.xls
93.18 KB
-
Phonic_AI_data_tabular.xlsx
14.66 KB
-
README_Fairbanks_et_al_2022_IJP.rtf
6.02 KB
-
us_corpus_compilation.xls
93.18 KB
Abstract
Language is a powerful form of communication that can reify and reproduce colonial legacies. For many primatologists – scholars who engage with diverse publics, ranging from personal social networks to formal classroom settings to myriad forms of science communication and outreach – it is common to encounter Anglophone speakers who add a final phoneme -ng, or /ŋ/, to the word “orangutan.” We interrogate and explicate the colonial and literary legacies of this phonological enigma. Structured as an essay, our article reports phonological survey results from 569 British- and North American-English speakers as well as a time series analyses sourced from Google Books Ngram Viewer. We found a large disparity between British- and North American-English speakers – 34% and 64% of which add the final /ŋ/, respectively – and telling reversals to the predicted extinction curve of “ourang outang” in Google Books’ British- and American-English corpora. Taken together, these findings put a new and problematic light on the final /ŋ/. Our intent is not to police the boundaries of acceptable discourse but to equip primatologists with the background and data needed for productively discussing and remedying colonial legacies during the course of educational and public outreach.
Methods
Figure 1 is based on an anonymous survey developed in Phonic AI (https://phonic.ai) and deployed globally using Mechanical Turk (https://www.mturk.com/) over a span of 3 months (March–May 2022). We obtained and evaluated audio recordings in response to a prompt—a photograph of an orangutan together with a written question, How do you pronounce the word orangutan?—from 569 respondents who self-identified as native Anglophones raised in North American or British commonwealth countries; however, 214 recordings (38%) were inscrutable and excluded from the analysis. We combined unambiguous responses from the United States (n = 175) and Canada (n = 52), because the English spoken in these two countries is more similar to each other than either is to British English (Boberg, 2010). The mean self-reported age of British- and North American-English speakers was 36.4 years (range: 19–77; n = 19) and 23.7 years (range: 18–49; n = 70), respectively.
Co-evolutionary dynamics between “ourang-outang” and “orang-utan” in the (A) British-English corpus and (B) American-English corpus (years: 1800 to 2019). Insets illustrate the exponential decay rate of “ourang-outang” between 1900 and 2019, with the exception of temporary reversals during the early 1930s and late 1960s. We downloaded data using the ‘ngramr’ package in R version 4.2.0 (Carmody, 2022; R Core Team, 2022) and followed best-practice guidelines (Younes & Reips, 2019); e.g., we standardized the frequencies by dividing them by the frequency of the word “an” in each corpus, which is indicative of total text volume, and subsequently taking z-scores of the resulting ratio.
Usage notes
The R Project for statistical computing: https://www.r-project.org/
The open source 'ngramr' package, freely available on CRAN: https://cran.r-project.org/web/packages/ngramr/index.html