There are basically three ways to deal with foreign names in English text.
- Preserve the original spelling of the name.
- If the name is originally written in the Latin alphabet, this can be done directly.
- If the name is originally in a different script, the equivalent is to use a transliteration convention that preserves orthographical distinctions present in the source language. For example, it's common to use
q in names of Arabic origin, even though English speakers will pronounce both as /k/.
- Preserve the original pronunciation of the name (to the extent allowed by English phonology).
- Translate the name into an English language cognate. For example, Spanish Juan → John, German Wilhelm → William, and Russian Екатерина → Catherine.
The transcription Юлия → Yulia results from option #2: Preserving the pronunciation /ˈjʉlʲɪjə/, or rather, its English approximation /ˈjuːli.ə/.
English orthography is famously phonetically ambiguous (e.g.,
a is pronounced /eɪ/, /æ/, /ɑː/, or /ə/ in different contexts), and thus it's not always obvious how to represent a sequence of sounds with letters (or vice-versa). But in general, English speakers have the expectation that
j is pronounced /dʒ/, and
y is pronounced /j/. And conversely, the expected spelling of /j/ is
y, and not
j. Hence, "Yulia".
Sure, some people may be aware of the fact that
j is pronounced /j/ in German (or Polish, Dutch, etc). But that's irrelevant. English speakers will read "Julia" as /ˈdʒuːli.ə/. Even if you expect your readers to recognize the name as being different from the English "Julia", why expect them to pronounce
J as German /j/ instead of as Spanish /x/~/h/, French /ʒ/, or Pinyin-Latinized Mandarin /tɕ/? We're translating from Russian into English here, so there's no point in introducing a spelling convention from some arbitrary third language.