Legal translation is often treated as a mechanical task. But in reality, it can have a profound impact on the outcome of disputes, particularly in international arbitration. A striking example is the ICSID case Kiliç v Turkmenistan, where the translation of a single clause shaped years of legal proceedings.
After reading Arsanjani and Reisman’s illuminating essay on the case in Practising Virtue [1], I revisited the underlying Decisions. What emerges is a textbook example of how high-stakes legal interpretation can hinge on translation—and how things can go wrong when linguistic nuance is overlooked.
The Clause at the Heart of the Dispute
The core issue revolved around Article VII.2 of the Russian version of the Turkey–Turkmenistan Bilateral Investment Treaty (BIT). Its English translation determined whether recourse to local courts was a condition precedent to arbitration before the ICSID. That question, in turn, had enormous implications for jurisdiction, procedural timing, and the investor's ability to seek relief.
Three Troubling Translation Issues
As a legal translator, several aspects of the case stood out:
1. A Lone, Anonymous Translator
The Respondent relied on a translation apparently performed by a single individual. The translation agency’s cover letter [2] notes: "we translated to the best of the translator’s knowledge and ability"—a telling phrase that distances the company from the actual work. No mention is made of the translator’s name, background, or legal expertise. Given the treaty's importance, this lack of transparency is troubling.
While a linguistic expert was later brought in to support the Respondent’s position, the original translator’s qualifications in Russian-English legal drafting were never addressed. That absence speaks volumes.
2. An Expert in Linguistics, But Not Legal Language
The Respondent’s linguistic expert, though competent in general English, was not trained in law. Her reasoning for omitting the word “if”—that two conditionals rarely appear together in standard English [3]—overlooks an important point: legal English routinely uses structures like “provided that, if...” as part of layered conditional clauses [4][5].
Arsanjani and Reisman rightly emphasize that legal language follows its own conventions. In this case, relying on a generalist rather than a legal-linguistic specialist may have undermined the analysis.
3. A Focus on English Grammar Over Russian Meaning
Perhaps most critically, the discussion centered on English grammatical coherence, not the meaning of the original Russian clause [6]. That’s a missed opportunity. Legal translation isn’t about producing grammatically pleasing English—it’s about conveying the precise legal intent of the original text. A bilingual legal expert would have been better positioned to assess this.
The Cost of Rushing Legal Translation
In closing arguments, counsel for the Claimant highlighted how much time and money could have been saved if proper attention had been paid to the multilingual versions of the BIT [7]. Even the Respondent admitted that “speed appears to have been a primary concern” during treaty drafting, leading to “accuracy” taking a back seat.
That admission captures a wider truth: legal translation is too often treated as an afterthought—outsourced to anonymous individuals under tight deadlines. Yet as this case shows, shortcuts taken during translation can end up consuming years of litigation and hundreds of thousands in legal fees.
Final Thoughts
Kiliç v Turkmenistan is more than an academic curiosity—it’s a cautionary tale. It reminds lawyers, arbitrators, and treaty drafters that translation is not a peripheral task. When stakes are high and texts are binding, the cost of imprecision is enormous. International procedural frameworks, such as the UNCITRAL arbitration regime, rest precisely on the fidelity of language versions. Investing in qualified legal translators from the outset isn’t a luxury—it’s a form of legal risk management.
-------------------------------------------------------------------------------------------------------------------------------------------------------
[1] Mahnoush H Arsanjani and W Michael Reisman, “Babel and BITs: Divergence Analysis and Authentication in the Unusual Decision of Kiliç v Turkmenistan,” in Practising Virtue: Inside International Arbitration, edited by David D. Caron, Stephan W. Schill, Abby Cohen Smutny, and Epaminontas E. Triantafilou, 407-425 Oxford: Oxford University Press, 2015.
[2] 4.20, Kiliç v Turkmenistan, ICSID CASE No. ARB/10/1, Decision on Article VII.2 of the Turkey-Turkmenistan Bilateral Investment Treaty, 7 May 2012.
[3] 9.15, Kiliç v Turkmenistan, ICSID CASE No. ARB/10/1, Decision on Article VII.2 of the Turkey-Turkmenistan Bilateral Investment Treaty, 7 May 2012.
[4] Mahnoush H Arsanjani and W Michael Reisman, “Babel and BITs: Divergence Analysis and Authentication in the Unusual Decision of Kiliç v Turkmenistan,” in Practising Virtue: Inside International Arbitration, edited by David D. Caron, Stephan W. Schill, Abby Cohen Smutny, and Epaminontas E. Triantafilou, 416. Oxford: Oxford University Press, 2015.
[5] Bryan A. Garner, Legal Writing in Plain English, 129. Chicago: University of Chicago Press, 2013
[6] 9.14, Kiliç v Turkmenistan, ICSID CASE No. ARB/10/1, Decision on Article VII.2 of the Turkey-Turkmenistan Bilateral Investment Treaty, 7 May 2012.
[7] 9.25, Kiliç v Turkmenistan, ICSID CASE No. ARB/10/1, Decision on Article VII.2 of the Turkey-Turkmenistan Bilateral Investment Treaty, 7 May 2012.
❓ FAQ: translation and the Kiliç v Turkmenistan arbitration
What translation issue was at the heart of Kiliç v Turkmenistan?
The dispute turned on how Article VII.2 of the Russian-language Turkey–Turkmenistan Bilateral Investment Treaty was rendered into English. The wording determined whether prior recourse to local courts was a precondition to ICSID arbitration, which in turn affected jurisdiction and procedural timing.
Why was relying on a single anonymous translator a concern?
The translation supporting one party came with a cover letter stating the work was done "to the best of the translator's knowledge and ability", without naming the translator or describing any legal training. For a binding treaty text, that absence of transparency about qualifications is significant.
Why does a general linguistics expert differ from a legal translator here?
The expert argued that two consecutive conditionals are unusual in standard English and so the word "if" could be omitted. That reasoning overlooks that legal English routinely uses layered conditional structures such as "provided that, if…". Legal drafting follows its own conventions, which a generalist may not weigh correctly.
What broader lesson does the case offer on legal translation?
It shows that treating translation as a peripheral, last-minute task can have lasting consequences. When binding multilingual texts are involved, imprecision can extend proceedings by years and raise costs considerably, which makes qualified legal translation a form of risk management rather than a luxury.
