Translation:Please give me two sheets of paper.
This is the funny thing about people asking about kanji. I can't see kanji when someone speaks Japanese. Thus, we shouldn't necessarily need kanji when Japanese is written. Plus, conextual clues and the actual subject of the study should clue people into the fact that no, we are not talking about gods...
I agree in principle, however I also find that Kanji make a written sentence much easier to parse (assuming of course that none are unknown to you). I can parse for example 新しい a lot quicker than あたらしい. This is especially true seeing as they make it easier to spot word borders in a script without spaces.
Moreover, missing kanji occasionally confuse the TTS engine. A few lessons on for example, there are sentences like 夕方はをみがきます。(I brush [my] teeth in the evening.) It is just very easy for the TTS engine to get this wrong and mistake は for the topic particle when in fact it is a noun here "tooth" (歯). And there are almost homophones which only differ in pitch accent, such as 橋 (bridge, はし) and 箸 (chopstick, also はし but with a high pitch on the first syllable).
It seems fine, I can understand it, but it's not what comes to mind (for me) when I think about individual sheets of paper. I've never heard anyone say 'paper sheets'. I live in Canada, so maybe it's just a regional thing. I would say 'sheets of paper' or 'pieces of paper'.
English uses counters too for certain things. "Two papers" would mean "two documents, two essays." In this case, though, you have to say "two sheets of paper." We also say "two loaves of bread" and not "two breads" (although people in Montreal often say "two breads," perhaps under the influence of French).
The pitch accent is different. In 神 (deity), the first syllable is high pitch, the second low (=> HL). Both 紙 (paper) as well as 髮 (hair) are LH.
According to what I hear (compounded by wikipedia), (Standard) Japanese words can be classified as having either one or no accent which is pronounced as a high pitch, followed by a downstep in pitch starting from the following mora (mora=syllable, but long vowels are two moras and -n is a mora on its own, too). A sequence of syllables without accent is a bit like a ladder with the pitch rising a little each. So if we use ꜜ to mark the place where the pitch drops (i.e. the preceding syllable is the accented one):
- Accent on the first mora (SꜜS...) has the first mora high and all following ones L, but with the ladder effect: HLLL... (this is the case in 神)
- Accent on any other mora (...SꜜS...): ladder effect on the mora before the ꜜ with the very last mora before ꜜ being the highest one, then the mora following ꜜ starts at the very bottom again.
- no accent: ladder effect throughout
So 神 is かꜜみ. For 髪 and 紙 I’m not quite sure if they are かみꜜ (accent on the last syllable) or accent-less – the ladder effect makes it nigh on impossible to distinguish them in isolation. Chances are at least one of them doesn’t have an accent because as far as I’m told the majority of native words don’t.
Also, note that since long vowels are two mora and ん counts as its own mora, the ꜜ can occur in the middle of a long vowel (e.g. 今日 きょꜜう) or between a vowel and -ん (e.g. 半 はꜜん). The result sounds like a falling pitch over the whole complex.
As far as I understand, you would sound odd but mostly understandable. Likely not quite as bad as if someone kept accenting the wrong syllables in English (probably because accent has a big influence on how the vowels are pronounced in English; compare for example how drastically different the vowel in the first syllable sounds in “photograph” vs. “photography”). There will also be occasions where a wrong pitch accent can be confusing on the listener (for example if you say はしに “at the end” (端に) when what you really mean is はしꜜに “at the bridge” (橋に)). In most cases context should resolve the issue, but be prepared for a bit of confusion here and there.