Hello/Bonjour Won’t Cut It in a Health Crisis - An analysis of language policy and translation strategy across Manitoban websites and social media during COVID-19 (2022)

This report provides an examination of language‑policy and translation strategy on Manitoban government websites and social‑media channels during the COVID‑19 pandemic.

Desjardins shows that multilingual health communication in Manitoba during COVID‑19 was uneven, hidden, and often inconsistent with both policy and demographic realities. By exposing these gaps, this article offers a clear roadmap for governments, NGOs, and scholars to redesign digital health messaging so that “Hello/Bonjour” truly cuts it in a crisis.

 What is this research about?

Goal: To understand how (or whether) multilingual communication was provided, how usable that communication was for non‑English/French speakers, and what the implications are for translational justice in a health crisis.

Guiding questions (implicit in the text):

• What translation‑related practices did the Government of Manitoba employ on its web‑presence and social‑media during COVID‑19?
• How accessible are those multilingual resources from a user‑experience (UX) perspective?
• What gaps exist between official language policy (English/French) and the linguistic reality of Manitoba’s population?

What do you need to know? – Context & Why It Matters

Background point	Why it matters for the study
COVID‑19 exposed language barriers – e.g., the Cargill High‑River meat‑packing outbreak where bulletin‑board notices were only in English, leading to confusion and higher case counts.	Shows the concrete health‑risk consequences of monolingual communication.
Canadian language legislation – Official Languages Act guarantees English/French only; Indigenous and migrant languages have no legal requirement.	Sets the policy ceiling against which provincial practice is measured.
Manitoba’s demographic reality – 2016 Census: ~22 % of residents do not have English as mother tongue; ~11 % speak a non‑official language at home.	Highlights a mismatch between population needs and the official‑language‑only approach.
Digital shift in crisis communication – Governments moved heavily to websites, Facebook, Twitter, YouTube; the speed of information delivery is crucial.	Provides the technological arena where translation (or its absence) becomes visible.
Research novelty – Uses a digital‑humanities toolbox (web‑scraping, hashtag indexing, network analysis, close‑reading) to study both the content and the UX of multilingual delivery, rather than a simple content‑comparison.	Offers a methodological blend rarely applied to language‑policy evaluation.

What did the researchers find? – Key Highlights & Supporting Quotes

Finding	Evidence (quoted)
Official‑language dominance persists – English and French dominate all COVID‑19 posts; other languages are rare.	“COVID‑19 Bulletin and the COVID‑19 Vaccine Bulletin are systematically posted in separate, language‑specific posts (English and French)… Neither … is translated or available in other languages.”
Multilingual resources exist but are hidden – Fact‑sheet translations (8 languages) and silent YouTube videos (12+ languages) are buried behind several clicks and not linked from the French site.	“Three mouse‑clicks from the English homepage will redirect a user to a section… where the Social (Physical) Distancing Factsheet can be found translated into eight languages.”
Inconsistent UX across language versions – The French site omits the multilingual fact‑sheets and video library that the English site shows.	“The French page does not have the seven other translated Social (Physical) Distancing Factsheets…”
Social‑media strategy is fragmented – Separate accounts for health officials, disjointed navigation to Shared Health Manitoba, and a predominance of English tweets.	“The Government of Manitoba’s Twitter account… tweets and retweets are predominantly in English, though French‑language content does make a somewhat regular appearance.”
Unequal YouTube playlists – English COVID‑19 playlist holds 346 videos; French playlist only 203, with far lower view counts.	“The English playlist total video count… is markedly larger than the French playlist… Engagement is also significantly lower for the French‑language playlist.”
Outlier success: Low‑German vaccine stickers – Community‑driven demand led to official production of stickers in Low German (Plattdeutsch), generating notable social‑media buzz.	“Andrew Unger … proposed … add Low German … and the Government of Manitoba obliged… the tweet garnered 283 likes, 28 quote tweets and 27 retweets.”
Platform omission – No active Instagram presence despite its potential for automated caption translation.	“The Government of Manitoba is oddly absent on Instagram… This seems like a missed opportunity…”

Especially Interesting Themes & Outlier Findings

• Hidden multilingual “gold mines”: The existence of multilingual PDFs and silent videos demonstrates that the government does produce content in many languages, yet the navigation design effectively hides it from most users.
• Language‑policy vs. UX paradox: Formal policies (e.g., French‑Language Services Policy) mandate bilingual service, but the user journey forces reliance on English‑dominant pathways, undermining the policy’s intent.
• Community‑led translation wins: The rapid adoption of Low‑German stickers shows that grassroots pressure can compel official bodies to expand language offerings beyond the statutory two.
• Social‑media fragmentation: Multiple official accounts (government, Shared Health, chief medical officer) dilute the information flow, making it harder for non‑English speakers to locate reliable updates.

How can you use this research? – Practical Take‑aways for Different Audiences

Audience	What to do (based on findings)
Public‑health officials / policymakers	• Audit all digital touch‑points (website, social media, video libraries) for discoverability of multilingual assets; redesign navigation to surface them from any language version.• Adopt a single, unified social‑media hub for health updates to avoid fragmentation.• Institutionalize a multilingual content checklist (incl. low‑resource languages identified in census data).
Government communications teams	• Implement language‑agnostic UI elements (e.g., language selector that stays visible on every page).• Publish parallel playlists on YouTube with identical video counts and promote them equally in both official languages.• Explore automated caption translation on Instagram and TikTok to reach younger, multilingual audiences.
Community organisations / NGOs	• Leverage existing multilingual PDFs/videos to create localized outreach kits (print, radio, community‑center displays).• Advocate for co‑creation of content with community members (e.g., the Low‑German sticker model).
Researchers / Academics	• Extend the digital‑humanities methodology (web‑scraping + UX analysis) to other provinces or sectors (education, emergency services).• Conduct user‑testing studies to quantify the impact of hidden multilingual resources on information uptake.
Future‑research agenda (as suggested by the author)	• Interview end‑users from Indigenous and migrant language groups to capture lived experiences of access barriers.• Compare Manitoba’s approach with provinces that have indigenous‑language mandates (e.g., Nunavut) to identify best practices.

What did the researchers do? – Methodology Overview

Methodological component	Details
Case‑study design	Focused on the Government of Manitoba’s digital communication during COVID‑19.
Data collection	• Web‑scraping of the provincial website and its “Resources and Links” pages.• Hashtag indexing / searching on Twitter and Facebook to capture COVID‑19‑related posts.• Network analysis of social‑media interactions (e.g., retweets, shares).
Qualitative analysis	Close‑reading of selected posts, PDFs, and video descriptions to identify translation cues (e.g., “translate” buttons, multilingual labels).
Quantitative snapshots	Counts of followers/subscribers (e.g., Facebook ≈ 58 k, YouTube ≈ 12 k); video counts per language playlist; citation of census statistics (language‑use percentages).
Ethical considerations	Only public‑facing data used; personal identifiers removed; compliance with Canadian Tri‑Council Policy on Human‑Subjects.
Scope limitations	No systematic surveys or interviews; analysis limited to publicly available digital artefacts up to mid‑2021.