Abstract
Past research has shown that national language policies can attract foreign direct investment (FDI), and that potential FDI-host countries coordinate their domestic language policies in anticipation of this. While the increased FDI-inflows arising from such language policies may benefit some members of society, these shifts in policy can adversely affect those whose spoken languages are not perceived as beneficial for attracting FDI inflows – primarily indigenous language speakers. This paper develops a theoretical framework to accordingly suggest that FDI inflows have contributed to declines in the usage and protection of indigenous languages in recent decades. This hypothesis is tested on a country-year sample of Latin American countries for the period 1988–2018. In evaluating this hypothesis with the aid of a newly constructed and comprehensive measure of time varying indigenous language usage spanning 20 Latin American countries and 30 years, FDI is determined to be a statistically significant contributor to the decline of indigenous language usage in contemporary Latin America.
Acknowledgements
The author of the paper would like to extend tremendous thanks to Benjamin E. Bagozzi for all of his input and help throughout this process. The author would also like to thank the anonymous reviewers of this paper, as well as the editorial board, for their excellent comments and suggestions. Furthermore, the author would like to thank Julio Carrion, Kyong-Min Son, and Christopher Wolcott.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 I define minority language as one that is not spoken by the majority of a nation’s population or one that is not protected as an official state language. This includes sign languages (University of Gothenburg, Citation2019).
2 I define government policy as a law, proposed agenda, or other government endorsement of an action that is directly related to an issue of public interest (University of Gothenburg, Citation2019).
3 Including water, soil, and hunting/farming land (José Quiroga, Citation2017; Jackson & Warren, Citation2005).
4 Obviously, there is wide discrepancy in the density of indigenous populations – 8% is merely an average. Mexico, Peru, Brazil, and Ecuador have some of the largest indigenous populations where Uruguay and Argentina have some of the smallest (Jackson & Warren, Citation2005).
5 Although the HME typology is a state-centric approach, there is merit as well to an individual-centric approach. One might argue that individuals choose their language in order to better than chances as jobs by foreign investors permeate the market. I do not discount that some individuals are indeed making this decision. However, I would argue that the state is first and foremost directing these macro-level language changes. The state is working with the MNCs, funding or defunding the indigenous language education, and directing far-reaching macro linguistic policy. Therefore, I argue in favor of a state-centric approach.
6 The MNCs are not always successful in these endeavors. However, under Schneider’s modification of the models of varieties of capitalism, hierarchical market economies do consistently seek to create a monopoly through these means, and this this is one of their primary motivating factors in investment in Latin America (see Schneider, Citation2013; and, for further information on hierarchy and economy, Back & Zavala, Citation2018).
7 Note that some FDI is indeed resource seeking. While such FDI may indeed be more focused in rural areas, my theory and mechanisms still hold even in this case. Moreover, even if market seeking FDI tends to locate in cities and touristic areas, its pressures on language policy will often affect all areas of a country since language policy is typically national-level—meaning that any adjustments to language policy via urban/touristic area FDI-pressures will shift language policy nation-wise. However, none of this should discount the nuance of FDI in the region.
8 Such as foreign portfolio investment.
9 For a full explanation of the politics of language, an excellent starting point is Coulmas (Citation2005).
10 By adopting the MNCs’ preferred language, indigenous peoples can be considered more qualified for promotion or more efficient for employment since the company does not need to invest in language training (Fredriksson et al., Citation2006).
11 Think of computer literacy or proficiency in a vocational skill.
12 See Appendix 1, Supplementary material for original text.
13 See Appendix 1, Supplementary material for original text.
14 The 1990s were crucial for indigenous populations in Latin America and elsewhere: After the UN began pushing for the protection of indigenous cultures worldwide, indigenous activists began petitioning their governments for recognition and protection (Jackson & Warren, Citation2005; McCarty, Citation2003).
15 The past literature that addresses routes through which minority language speakers have attempted to protect and use their indigenous languages outside of education is relatively new. It primarily focuses on the regime types (i.e., democracies’ treatments of minority languages versus authoritarians’ treatments of minority languages) (Safran & Liu, Citation2012; Liu, Citation2011; Liu et al., Citation2016) and/or indigenous language use in government settings.
16 That this author can find as of 2019
17 As the other databases do. The Soviet dataset is particularly guilty of this (Anon, Citation1964). Of course, these datasets all serve purposes in their own right, and are not to be generally disparaged.
18 Although there are instances where SIL is the only organization to have measured a specific language.
19 Also including Mexico and excluding French Guiana.
20 For a fuller exploration into my usage of Ethnologue, see Appendix 2
21 Many languages with missing data had fewer than fifty speakers
22 I assumed that a family was made up of 4 members – this value will not affect my data analysis since I am only looking at languages with more than 50 speakers
23 For a full list of control variables, see Appendix 3, Supplementary material.
24 In both cases, the averaged values have been rescaled for interpretability by dividing each raw average by 1,000,000.
25 For the full tables of coefficient estimates and standard errors, see Appendix 4, Supplementary material.
Additional information
Funding
Notes on contributors
Sarah A. V. Ellington
Sarah A. V. Ellington is a PhD candidate at the University of Delaware’s Department of Political Science and International Relations in Newark, DE, 19716 (contact: [email protected]). Her research focus is on indigenous language erosion and revitalization efforts in Latin America.