English vs. Devanagari Trademarks: Script and Transliteration Rules

The legal test is the customer, not the alphabet

Section 18(1) refuses marks that are identical or deceptively similar — and deception is assessed from the ordinary consumer's point of view. In Nepal that consumer navigates two scripts daily: a signboard can read EVEREST, एभरेष्ट, or both stacked. If the two marks would be pronounced the same or recognised as the same name, the script they are written in does not save the later filer.

Equally, script is not irrelevant: a Devanagari mark and a Latin mark with genuinely different sounds and appearances can coexist. The question is always whether the ordinary buyer, hearing or seeing both, would think one brand.

• Similarity is assessed by sound and commercial impression — script is a wrapper, not the test
• Transliteration equivalence: एभरेष्ट ↔ EVEREST is one name in two clothes
• Genuinely different-sounding marks in different scripts can coexist

One script, one filing: the coverage question

A registration covers the mark as filed. A Latin-script registration does not, by itself, guarantee protection against a Devanagari copy — and vice versa. For a brand that trades in both scripts (which in Nepal is most brands), the strategic options are:

• File both forms as separate applications — the cleanest coverage, at twice the applications (still one class each under Section 18A)
• File a composite mark — Latin and Devanagari in one label — protecting the combination, though less cleanly each element alone
• File the primary trading form and monitor the other script through a watch — pragmatic, with a documented residual risk

Transliteration is a search problem before it is a legal one

The practical difficulty is finding the conflicts: Devanagari-to-Latin transliteration is not one-to-one. एभरेष्ट could be romanised EVEREST, EBREST, ABHIREST or worse; SAGARMATHA could appear as SAGARMATHA, SAGAR MATHA, or with the aspirated variants Nepali spelling allows. A search that only matches literal strings will systematically miss the mark that is, commercially, yours.

Professional screening solves this with variant generation: for each significant word, generate the phonetic and transliteration family — including the PH/F, C/K, S/SH, doubled-consonant and vowel-length families — and search them all, then let class and goods filtering sort the noise. This is exactly what the tools on this site do across 71,000+ published records.

Drafting and examining bilingual applications

The Department accepts applications in Nepali or English, and examination correspondence follows accordingly. For a bilingual brand, consistency matters: the transliteration you file should be the transliteration you trade under — a mark registered as एभरेष्ट while the signage says EBHAREST invites both customer confusion and enforcement friction.

For foreign brands, the decision is sharper: enter in Latin script only, or localise. Localisation builds consumer connection but creates a second mark to clear, file and renew — usually worth it only once the Latin-script mark is secured and a watch is running.

• File the transliteration you actually trade under — register and signage should agree
• Foreign entrants: secure the Latin form first; localise deliberately, not by accident
• Watch both scripts regardless of what you filed — the conflicts arrive in both

The watch requirement that follows

Because conflicts cross scripts, post-filing monitoring must search both scripts too. A watch limited to the script you filed in leaves the एभरेष्ट-for-EVEREST application unopposed until it registers — the exact failure bilingual clearance exists to prevent.

Script-aware watching, with transliteration-aware matching and BS/AD deadline computation, is the monitoring standard a bilingual market actually requires.