The standard recommendation and its blind spot
There’s one rule almost everyone in this industry agrees on: a translator should work into their mother tongue. ISO 17100, the dominant European quality standard, builds the buyer’s core guarantee around exactly that idea—work completed by a professional translator into his/her native language,
checked by a reviewer who also shares this same mother tongue.
The UK’s Institute of Translation and Interpreting writes it into its code of conduct. It’s sound advice, and I follow it.
But notice what the rule is silent about: whether the translator actually understands the source. For a birthday card or a product blurb, that’s a fair assumption. For a DAX-adjacent annual report—where a single German sentence can carry a board’s carefully calibrated reassurance—it isn’t. I’ve watched fluent English translators produce prose that reads beautifully and quietly misstates what the German meant. The mother-tongue principle protects the output and says nothing about the input.
So the honest version of the rule has two halves. You need native-grade intuition in the target language and native-grade comprehension in the source. Those are two separate competences, and most working translators hold one firmly and the other by approximation. The only way I know to hold both at full strength at once is to have grown up in both languages. That’s the case I want to make—carefully, because bilingualism alone doesn’t finish the argument either.
What growing up bilingual actually means—and why it’s rare in this work
There’s a real difference between simultaneous childhood bilingualism and learned near-native fluency, and it matters here. A gifted L2 speaker builds a second language on top of a first: one system is home, the other is modelled—brilliantly, sometimes—against it. Someone who acquired both languages from birth has two home systems. Not one native intuition and one excellent reconstruction, but two separate native intuitions, each with its own reflexes for register, rhythm, and what simply sounds wrong.
For the record, that’s my own background: German father, American mother, both languages from the start. It’s why I can read a Vorstandsvorsitzender’s letter and feel the German subtext, then hear whether the English draft lands the same way—two reflexes, not one reflex and one rulebook.
I’ll resist overselling it, though, because the research won’t let me. Translation scholars are blunt that bilingualism is necessary, but not sufficient for translation proficiency and efficiency
—and that a born bilingual often suffers from not truly knowing any language well enough to translate.
Two native intuitions are the raw material. What turns them into a usable annual-report translation is years of specialist practice on top. Bilingualism buys you the ear; the reporting seasons teach you what to do with it.
The underrated half: why source-language depth counts just as much in a report
German corporate prose is a coded register, and the code carries money. When a Vorstandsvorsitzender writes of a solides Fundament für nachhaltiges Wachstum,
that isn’t decoration—it’s a signal calibrated for analysts. Planmäßiger Verlauf
is quiet reassurance that nothing derailed. An HGB ausgeglichenes Ergebnis
is strategic understatement, not an accounting shrug. A translator who parses these correctly but doesn’t feel the rhetorical intent will produce English that is accurate word by word and wrong in effect.
This is where the mother-tongue principle fails silently. The most striking evidence I’ve seen comes from the head of a large state institution’s translation department, who explained that they didn’t use English native-speaking translators for into-English work because they didn’t understand the Dutch source texts well enough.
The unit chose source comprehension over target nativeness on purpose. Legal-translation scholarship has started arguing the same thing for dense texts: the ideal translator, as one scholar has put it, might actually be a native-speaking (legal) expert of the source language
able to explain the peculiarities of the issue in full.
A small, real example of the trap. German management commentary: Der Vorstand geht davon aus, dass sich das Ergebnis im weiteren Jahresverlauf planmäßig entwickelt.
The literal-but-deaf rendering: The Management Board assumes that earnings will develop according to plan over the remainder of the year.
Grammatically fine. But planmäßig
here is a confidence signal, and assumes
undercuts it—in English, assumes
leans toward guesswork. What the board actually conveys is closer to: The Management Board expects earnings to remain on track through year-end.
Same facts, restored reassurance. You only make that call if you heard the German first.
Terminology hides the same depth problem. As one German corporate-language agency notes, Abschreibungen
is possibly the single most challenging term in a financial report—because it maps to depreciation, amortisation, or impairment depending on context you can only read from the German. No glossary lookup closes that gap; comprehension does.
Target-language quality: what real business English does that translated English doesn’t
Now the other half. The most visible fault line between German and English is structural. Formal German, as grammar references put it plainly, prefers to pack actions into nouns rather than spell them out as verbs
—the famous Nominalstil. English plain-language convention runs the opposite way; generations of native writers have been trained by Strunk & White, Orwell’s rule, and the Plain English movement to avoid nominalizations.
A German-dominant translator will often carry the nominal structure across intact. It’s not ungrammatical. It just reads as translated.
So the realization of synergies
survives where a native writer reaches for capturing synergies.
The implementation of the strategy
stands where an English ear wants rolling out the strategy.
Multiply that across a forty-page chairman’s statement and the prose acquires a uniform, slightly airless weight—correct, joyless, unmistakably converted from something else.
Underneath structure sits texture: collocation. Peer-reviewed research is consistent that L2 speakers differ from natives in their collocations both quantitatively in terms of the number and types of collocations used, as well as qualitatively in terms of error-free use
—a gap that persists at advanced proficiency. Worse for translators, L2 speakers tend to rely on using L1 translation equivalents
: the German shapes the English choice, toward the congruent option that’s often a shade off. A controlled psycholinguistic study found native English speakers were sensitive to collocation restriction, whereas the non-native speakers were not
—and, revealingly, the non-natives were the ones sensitive to L1–L2 congruency. That’s the mechanism, measured: the native rejects the not-quite-right pairing on reflex; the L2 speaker is pulled toward it by the source.
This is exactly where report-level judgment lives. In IFRS notes and GRI/ESRS disclosures, a lot of the text is template—defined terms, standard tabular language—and there a native ear adds little. But the moment you cross from template into the client’s narrative voice—the discourse markers that stitch a long chairman’s statement together, the cohesion across a sustainability section, the cadence of an investor letter—you’re writing English that has to sound authored, not assembled. The market has quietly conceded the point: serious annual-report shops now only entrust our texts to native speakers of the target language
with financial-publication experience, aiming for a translation that reads as though it were written in the target language.
Where even good L2 translators hit systematic limits
I want to be fair here, because this isn’t a competence insult. I work alongside L2-English translators who are sharper on German tax law than I’ll ever be, and their drafts are clean. The limits I’m describing are systematic, not personal—they sit in the parts of language that operate below conscious rules.
The first is collocational plausibility. An L2 draft can be fully correct and still trigger mild unease in a native financial analyst—a pairing that’s grammatical and attested but not what the register reaches for. The non-native reader won’t notice. The native buy-side reader will, without being able to say why. That’s the collocation-restriction sensitivity the psycholinguistic work documents, playing out in a real audience.
The second is register drift in sustainability narrative. A great deal of ESG English written on the continent sounds European-institutional—Brussels-report cadence—rather than London-boardroom. It’s a perfectly respectable dialect of English, but it isn’t the voice a UK or US investor expects from a company addressing them directly, and the mismatch reads as distance.
The third is the hardest to pin down: what native English readers often describe as a subtle unevenness they can feel but can’t name—a faint fog of slightly unidiomatic prose across an otherwise flawless audit-context document. No single sentence is wrong. The cumulative texture just isn’t native. In a document whose entire job is to project control and credibility, that low-grade static is not a cosmetic problem.
When bilingualism is decisive—and when MTPE closes the gap
None of this means every line of a report needs a bilingual native. Honesty about stakes is part of the service, and the report itself tells you where they’re high. The templated machinery—pension provisions, deferred-tax reconciliations, lease tables, standard IFRS and HGB notes—is highly formulaic, heavy with repetition, and well served by translation-memory leverage and supervised machine-translation post-editing (MTPE). The English there is a known quantity; native intuition adds little a good TM and a careful post-edit don’t.
The narrative sections are the opposite. The chairman’s statement, management commentary, sustainability storytelling, the CFO or investor letter—these are where register and credibility signalling are the meaning, and where native bilingual intuition is mission-critical. In my own MTPE work on a CEO letter from raw machine output, I keep maybe one sentence in five; the rest I rewrite, because tone is the message and the machine can’t hear it.
17 years of practice: what DAX-adjacent reports taught me about linguistic authenticity
I’ve been translating full-time since 2008, north of 7,000 jobs, a lot of them reporting-season work under NDA. A few patterns repeat every year, and they’re what turned my instinct on this into a firm position.
There’s the CFO letter that works beautifully in German—measured, quietly confident—whose faithful English translation somehow reads like a press release from a different, brasher company. Nothing is mistranslated. The register has simply migrated, sentence by sentence, until the voice belongs to someone else. Restoring it isn’t correction; it’s re-hearing the German and re-speaking it.
There’s the sustainability section a client’s English-speaking IR officer sent back with three words in the margin: this doesn’t sound like us.
She was right, and she couldn’t have pointed to a single broken sentence. That’s the fog—cumulative register drift that only a native reader downstream detects, at exactly the moment it’s most expensive to fix.
And there was a sentence, in a recent season, that needed three native-intuition decisions in a row: a Nominalstil construction to unwind into English verbs, a German-congruent collocation to reject in favour of the one the register actually wants, and a hedged German forecast verb to pitch at the right confidence level for an analyst audience. Any one of the three, an excellent L2 translator makes correctly. All three at once, on reflex, in the time it takes to type the line—that’s what two native systems plus seventeen years buys. Not magic. Just two ears where most of the work has one.
What this means for your next project
When you shortlist a vendor, test genuine bilingualism instead of taking native fluency
on faith. Three questions do most of the work. First: which languages did they acquire in childhood, and can they say so specifically? Second: how do they handle a hedged German forecast verb like dürfte
—do they translate it or calibrate it? Third: ask for a short sample from real management commentary, not a clean marketing text.
In that sample, read for three things: register consistency across the whole passage, collocation patterns that sound authored rather than assembled, and a healthy nominal-to-verbal shift—German nouns turned into English verbs where the plain-language norm expects it.
A translator who grew up with both German and English doesn’t choose a register—he hears it. That’s the difference between an annual report that sounds written in English and one that sounds translated into it.
If your next report has a chairman’s statement, a CFO letter, or a sustainability narrative that has to sound like your company in English, that’s the work I do. Get in touch, or see how I price it on the services page. For a straight terminology question, I’m happy to just answer it.