> I noticed some typos in https://sqlite.org/whentouse.html that I wanted to fix. Firstly to your question, some very staunch grammar hawkeyes on here correct documentation errors all the time, you are welcome to post any needed corrections right here, it often will be fixed within the day. Secondly, that use of "an" is perfectly correct. The English rule is that a word starting with a *vowel sound* (not necessarily a vowel) gets the "an" rather than the "a" (I will spare everyone a dissertation on why that is, but it's easy to look up). That means words like "Hour" and "Yttrium" gets the "an" even though they do not start with strictly a vowel but have a vowel sound, similarly some vowel-starting words like "Union" (starting with j sound same as Yellow) also gets an "a" rather than an "an", etc. Which brings us to acronyms. Typically if an acronym (or even just a letter) is pronounced by wording it's name (M = Emm, S = Ess, and others like F, L, N, R, X) all gets an "an", whereas other letter names like B, C, D, U, Z all do not start with a vowel sound and so only gets the "a". So far so easy, but then we get to specifically "SQL", it's a little bit more tricky because the phenomenon called "SQL" is often pronounced two different ways by people, both of them acceptable pronunciations: some say "Ess-Queue-Ell" and others say "Sequel". Sounding out those words makes it obvious that one should get an "a" and one should get an "an". Thus, it is not only "acceptable" linguistically to use both, I propose that it is very helpful to let the reader know which pronunciation the writer is intending when they write "SQL" in a sentence.