Thanks to lexicographers

[Joanne M. Despres]

I should make it clear that my staff DOES try to distinguish
between citations of foreignisms in English-language context and
truly naturalized usages.  We rely on cues such as quotation
marks, italicization, or introductory language such as "as the
French say..." or "as it's called in Greek..," etc.  It's not always
easy to make the distinction, of couse, but we certainly do our
best.

Translated text or writings dealing with exotic cultural phenomena
(such as ethnic cuisine) offer both rich and potentially  treacherous
material for first-occurrence-hunting.  On the one hand, you're more
likely to get very close to an absolute first use in a context like that
than almost anyplace else; on the other hand, you're also very
likely to be dealing with words that aren't yet truly established in
the language.  In cases like these, one's judgment about whether
the occurrence in question is truly the direct descendant of the
entered word (that is, whether it represents the earliest known
example of a history of continuous use) depends to a large extent
on (apart from the red flags previously mentioned) whether the word
happens to be attested again fairly soon after the apparent original
use.  But of course, it's impossible to know for sure  whether a
period of non-attestation means the word wasn't actually in use, or
was rarely used, or simply happened not to be recorded.  At M-W
we have a tendency (as in so many things) to err on the side of
caution and disregard isolated early occurrences in deciding on the
date of a word, but those early borderline cases are definitely an
important part of the linguistic record, and I'm very glad the OED
reports them.

Joanne Despres
Merriam-Webster