The Bilingual
Disclosure Handbook
Why the English version of your Thai ESG report decides your score — and the terminology, traps and QA that keep the two languages saying exactly the same thing.
The English version of a Thai ESG report decides its FTSE, MSCI and CDP score; the Thai 56-1 One Report is what legally prevails. When the two drift — a hedged Thai “มุ่งสู่” hardened to English “will”, carbon-neutral swapped for net-zero, a Buddhist-Era year copied raw — the company loses rating points or gains greenwashing exposure (the US SEC fined Invesco US$17.5m over ESG wording; the EU regime bites from Sep 2026). The fix is procedural: one locked bilingual termbase, ISO 17100 translate–edit–proofread, and an MQM-scored claim-parity check — this handbook is that procedure.
The two-language reality
Your Thai report is filed for the regulator. Your English report is scored by the market. When the two drift apart, you lose the points the Thai version earned.
Global ESG assessors read the English. FTSE Russell scores companies from their own published disclosures, and CDP is explicit that scores are "based solely on the data provided in your response," reflecting "the quality, completeness and maturity" of the disclosure — not the underlying performance. To a London or Singapore analyst working from your English 56-1, a disclosure that is clear in Thai but garbled in English is, for scoring purposes, undisclosed.
Thai regulators file the Thai. The 56-1 One Report is filed in both languages, but the Thai version is the legal submission and prevails in any Thai proceeding; the English is an official translation for investors. So a Thai issuer runs two documents that must carry exactly the same meaning — one that governs legally, one that governs your rating. And the research is clear that disclosure clarity, not just quantity, drives rater agreement: ambiguity "leads to varied interpretations and distortions by rating agencies." A weak English rendering re-injects the very ambiguity that pulls your scores apart.
The terminology that must reconcile
A handful of ESG terms carry precise, scored meaning. Rendered loosely — or differently in different places — they quietly cost points. These are the pairs a locked bilingual glossary exists to fix.
| Thai term | English (preferred) | Common wrong rendering | Why it moves the score |
|---|---|---|---|
| ก๊าซเรือนกระจก | Greenhouse gas (GHG) | "carbon" | Collapsing GHG into "carbon" understates CH₄/N₂O and breaks GHG-Protocol / CO₂e consistency. |
| ความเป็นกลางทางคาร์บอน | Carbon neutrality | "net zero" | Carbon-neutral = offsetting today's emissions; equating it with net zero is a top greenwashing flag. |
| การปล่อย…สุทธิเป็นศูนย์ | Net zero | "carbon neutral / zero carbon" | Net zero = ~90% abatement across Scopes, removals only for residuals; raters score the pathway, not the label. |
| สาระสำคัญ / ประเด็นสำคัญ | Materiality / material topics | "important issues" | "Material" is a defined reporting threshold (GRI/ISSB); "important issues" removes the assurance hook. |
| การทวนสอบ | Verification | "assurance / audit" | Verification (ISO 14064-3) ≠ assurance (ISAE 3000); swapping them overstates the credibility level. |
| การรับรอง | Assurance | "verification / guarantee" | Assurance carries a limited-vs-reasonable grade; "guarantee" implies a promise no assurer gives. |
| ขอบเขตที่ 1 / 2 / 3 | Scope 1 / 2 / 3 | "level 1" / "supply-chain emissions" | GHG-Protocol fixed terms; Scope 3 ≠ "supply chain" — 15 categories, ~70–90% of the footprint. |
| การชดเชยคาร์บอน | Carbon offset | "carbon reduction" | Offsetting (buying credits) is not reduction; SBTi bars offsets from near-term targets. |
| เป้าหมาย | Target | "goal / aim / aspiration" | A target implies a quantified, dated commitment; analysts score targets and discount aspirations. |
| มุ่งสู่ / มุ่งมั่นที่จะ | Aims to / working toward | "will / commits to" | Over-hardening an aspirational Thai verb into "will" manufactures a commitment never made. |
| คาร์บอนฟุตพริ้นท์ขององค์กร | Carbon Footprint for Organization (CFO) | "CFO (Chief Financial Officer)" | In Thai ESG, CFO = the TGO organizational-footprint scheme — spell it out to avoid the finance-role clash. |
| ผู้มีส่วนได้เสีย | Stakeholder(s) | "shareholders" | Stakeholder ≠ shareholder (ผู้ถือหุ้น); conflating the two misreports the materiality-engagement base. |
| การขับเคลื่อนธุรกิจเพื่อความยั่งยืน | Driving business toward sustainability | "CSR" | The SEC deliberately renamed the CSR section; keeping "CSR" signals an outdated, philanthropy-framed report. |
One term alone appears three ways in Thai reports — สารัตถภาพ, สาระสำคัญ and ประเด็นสำคัญ all for "materiality." A per-translator choice makes the same concept look like three; a locked glossary maps every variant to one English target, every time.
The mistranslation traps
Some errors don't just read awkwardly — they change a rating-relevant fact. These are the top traps, ranked by consequence.
- 01Carbon-neutral ⇄ Net zero — Swaps a weak, offset-based claim for a science-based one (or vice versa) — the top greenwashing trigger and a direct FTSE/CSA point swing.
- 02Verification ⇄ Assurance — การทวนสอบ vs การรับรอง — overstates the third-party credibility level; assurance carries a limited/reasonable grade verification does not.
- 03Offset ⇄ Reduction — ชดเชย vs ลด — presents bought credits as real abatement; SBTi and raters score them oppositely.
- 04Target ⇄ Aim / aspiration — เป้าหมาย vs มุ่งสู่ — turns a soft ambition into a hard, dated commitment (liability), or dilutes a real target (score loss).
- 05"มุ่งสู่" hardened to "will" — Manufactures a promise the board never approved — reverse-greenwashing and misleading-disclosure exposure.
- 06Scope 2 "indirect", unqualified — Blurs Scope 2 vs 3 and hides location- vs market-based method, hurting data-completeness scoring.
- 07Material topics → "important issues" — Strips the defined reporting threshold, weakening the link to assurance and double-materiality logic.
- 08Scope 3 → "supply-chain emissions" — Silently drops downstream and most of the 15 categories — a completeness penalty.
- 09CFO read as Chief Financial Officer — An acronym collision unique to the Thai TGO scheme — comic in a boardroom, costly in a filing.
- 10Buddhist-Era date copied raw — 2568 (พ.ศ.) is 2025 (CE). Copied straight into English, that's a 543-year error in a target year.
The claim-strength trap
English is denser in green-marketing intensifiers than Thai. A cautious Thai phrase can drift into an absolute English claim — the exact trigger regulators act on, regardless of intent.
Greenwashing enforcement targets the wording, not the intention: "regulators do not require intent — a claim that misleads is actionable regardless of whether the company meant to mislead." A translation that upgrades a hedged Thai statement into an unqualified English one manufactures a claim the company never made.
And the exposure is cross-border. The US SEC fined Invesco US$17.5m and DWS US$19m for misleading ESG statements; the EU's Empowering Consumers Directive applies strict anti-greenwashing rules from September 2026; and Thailand's 56-1 One Report mandates ESG disclosure on a comply-or-explain basis. A Thai issuer's English report is read in all three. The QA job is claim-strength parity: the English must be neither stronger nor weaker than the substantiated Thai source.
| Cautious source intent | Over-claiming drift (risk) | Accurate, qualified English |
|---|---|---|
| "มุ่งสู่ความเป็นกลางทางคาร์บอน" (working toward carbon neutrality) | "We are carbon neutral" | "We have committed to a carbon-neutrality target by 2050" |
| "เป็นมิตรต่อสิ่งแวดล้อม" (environmentally considerate) | "100% eco-friendly / green" | "Reduces [impact] by X% vs [baseline], verified by [assurer]" |
| reduction "ตามเป้าหมายภายใน" (per internal target) | "SBTi-aligned science-based target" | "Internal reduction target; not yet externally validated" |
Why Thai makes this hard
Thai isn't a harder language — it's a differently-shaped one. Several of its features force the translator to make rating-relevant choices English speakers never see.
- ◆No verb tense or future marker. Thai marks the future only with จะ. A target sentence without it is ambiguous between "reduces" and "will reduce" — the translator chooses the tense, and that choice decides whether a claim reads as achieved, ongoing or promised.
- ◆Soft, gradable modality. มุ่งมั่น / มุ่งสู่ / ตั้งเป้า / คาดว่า span aspire → aim → target → expect. English forces a commitment level; "will" vs "aims to" is a rating decision, not a neutral swap.
- ◆No plurals or articles. "โรงงาน" is one plant or every plant — in an emissions boundary, that difference changes your reporting scope.
- ◆Loanword vs coinage drift. The same concept appears as a transliteration (คาร์บอนนิวทรัล) and a Thai coinage (ความเป็นกลางทางคาร์บอน) in one report — a locked glossary maps both to one English target.
- ◆Buddhist-Era dates. Reports may use พ.ศ. (CE + 543); 2568 is 2025. Copied straight, that is a 543-year error in a target year.
- ◆Run-on clauses & meaning-bearing spaces. Thai chains clauses without full stops and has no word spaces; mis-segmentation around numbers and units can flip meaning.
- ◆Formal, agentless register. 56-1 Thai is elegant, nominalised and often omits the actor; English governance and assurance text usually needs the responsible party named.
The QA layer that catches it
Translation quality is part of the disclosure control environment — and it can be measured. This is the workflow a procurement-grade bilingual disclosure runs.
- 01Termbase governance. A concept-oriented, client-approved ESG glossary (Scope 1/2/3, net zero, materiality, GRI / ISSB / TGO terms) locked across the whole report and year on year.
- 02ISO 17100 translate → edit → proofread. An independent second linguist compares target to source for accuracy, completeness and terminology; then an ESG subject-matter review; then proofread.
- 03Back-translation & reconciliation. The highest-risk passages — targets, claims, forward-looking statements — re-translated to Thai and reconciled. The pharma gold standard, applied to material ESG claims.
- 04Claim-strength parity check. An explicit review that no claim reads stronger or weaker than the substantiated source.
- 05Numeric & format integrity. Every figure verified against source (tCO₂e, %, baselines, target years, BE↔CE dates, separators); totals reconcile across narrative, table and appendix.
- 06MQM-scored QA report. Errors tagged by type and severity, aggregated into an auditable score with a pass threshold.
- 07Sign-off & provenance. Who translated, revised, reviewed and approved; termbase and source versions; the QA score — a trail that files alongside your assurance and audit evidence.
| Error category | What it catches | Why it's high-stakes in ESG |
|---|---|---|
| Accuracy | Mistranslation, omission, addition | A dropped clause is an omitted disclosure; carries the claim-strength and numeric risks. |
| Terminology | Wrong or inconsistent term vs the glossary | "Net zero", "Scope 3", "assurance" carry regulatory meaning; inconsistency reads as unreliability. |
| Fluency | Grammar, spelling, coherence, register | Poor fluency signals low control and erodes analyst confidence in the whole report. |
| Style | Awkward, over-intensified, register mismatch | Marketing-flavoured English inflates claims — straight into greenwashing exposure. |
| Locale | Number, date, currency, unit formatting | THB/USD, BE↔CE dates, tCO₂e, separators — a formatting slip becomes a factual error. |
How Othello does it
This is Practice II — Disclosure Linguistics — the bilingual QA layer no generic language vendor runs, powered by our ESG Disclosure Studio.
Send us your report — Thai, English, or both.
We’ll run an indicative bilingual disclosure QA — terminology, claim-strength parity and numeric consistency — and show you exactly where the two languages diverge. Within the hour, under NDA.