[LI] Languages & UBI; lang tech in Africa; MT; missing words

Four items posted to LinkedIn relating to languages, one with economics and three with technology.

Question about how UBI might affect language usage (posted September 2019)

How might universal basic income (#UBI) affect use of various #languages? Question arises from hearing again the notions “can’t get a #job speaking X” & “speaking Y improves job prospects.” Asking not as an advocate of UBI, but am following the topic

(my response to a comment re Bourdieu’s notions of prestige language and linguistic capital)

Thanks for all the suggestions. Might also have to look at other work on the economics of language. Not sure where I’d find the time, but would be interesting to explore.

(my response to comment re languages being entangled)

True, I guess, but this question prompted by a phrase in a TED talk that reminded me of “Who’s going to hire you if you speak Zulu?” from a conversation regarding language in South Africa (of which I forget the specific context). It’s quite common, I believe, for parents to encourage their children to learn (in) a dominant (usually Europhone) language to the extent even of discouraging their using & learning their maternal language, based on the notion of potential economic return.

The major fallacy in this reasoning, of course, is that one doesn’t have to *not* learn one language in order to gain/maintain competence in another. However it is hard to deny that speaking dominant/prestige language at least leaves open some potential lines of employment (given skills, connections, etc.).

So I was just wondering whether a guaranteed income (#UBI) might mitigate the seeming all-or-nothing quality of the economic argument for learning “Y” & not “X.” Not aware of anyone having gone there…

(my response to comment about major languages and conflation of questions)

Belated reply… Yes there is the angle of economic value of having a “dominant” language. I don’t see the original question as conflation, but rather looking for secondary effects, Put another way, might UBI lessen the tendency to see language only in terms of economic return, and hence in effect leave more space for minoritized languages?

Repost with comment of Donna Parrish’s post of an article entitled “Ranking Language Technology Regions NIMDZI, October 2019 (posted October 2019)

Re Africa as “the land of localization opportunity”: “58 percent of respondents in North Africa, and nearly 40 percent of respondents in Sub-Saharan Africa reported that the internet’s lack of culturally and linguistically-relevant content was one of the most significant reasons why these mobile users chose not to access the Internet.”

With its rich linguistic diversity, couldn’t Africa be a leading market for & source of innovation in language technology?

Recent advances in low-resource machine translation,” Facebook Artificial Intelligence, by Marc’Aurelio Ranzato, Paco Guzmán, Vishrav Chaudhary, Peng-Jen Chen, Jiajun Shen, and Xian Li, 16 October 2019 (posted October 2019)

One of my hopes for machine translation (MT) 2 decades ago was that it might soon work to the advantage of many non-dominant languages (which in much of the world are “less-resourced”) – such as those of Africa. That road has been long, however, especially as MT was taken on an important detour through statistical methods that require large corpora. This research, however, seems to indicate that a corner may have been turned.

Reply to repost by Donna Parrish of a post by Translators Without Borders of an article entitled “What happens when 12m people have no word for ‘dinosaur’ (BBC, 16 January 2020) (posted January 2020)

Brings to mind this 2014 piece on translating technical terms for software localization, which highlights Ibrahima Sarr’s work on Fula for Mozilla: https://www.economist.com/international/2014/09/27/cookies-caches-and-cows


Other blogs > LinkedIn > LinkedIn articles & posts, 2020

Leave a Reply

Your email address will not be published. Required fields are marked *


This site uses Akismet to reduce spam. Learn how your comment data is processed.