This is, technically, a good point that Japanese is not perfectly 1-to-1 writing-to-pronunciation in the kana syllabaries (I guess in particular Hiragana, as I imagine it’s probably rare to write actual Japanese particles like the topic particle as Katakana (ハ). And, it is commonly claimed (even in some textbooks; even when they go on, in the next few pages to contradict themselves) that learning the ‘defined’ Hiragana sounds allows you to correctly read all Japanese written in Hiragana.
Exceptions do indeed exist.
On the other hand, the way you wrote it, you were seemingly claiming that it was common with Kanji for the official reading to be different from the official pronunciation. I don’t have hardly any experience with Japanese-in-the-wild yet, only really from learning resources (will remedy that soon, but that’s the situation for the moment), but I haven’t run across any examples I can remember, off the top of my head, where a Japanese word’s given reading is different from how it’s actually pronounced.
Because, although it’s natural to assume that kanji have prescribed/strictly-official readings (and I even basically made it sound like that above), my understanding has shifted a bit recently, and it seems to me more accurate to say that pronunciations are more tightly associated with words than kanji, and the kanji guide the typical pronunciation of the word, but some words (for example) started with some particular pronunciation, and the kanji was associated with the word later on.
Sometimes that would mean that the kanji would pick up a ‘new reading’, like I imagine how they picked up the kunyomi readings. But there are some words, even very common words, where the kanji is used purely for meaning and not pronunciation, and there are even words where although it the kanji seems totally natural and seems like it should define the pronunciation, nevertheless the word has an unrelated pronunciation.
Yet – and this is my main point – all this really means to me is that the pronunciation is more associated with the word itself. And the kanji, while often being used to guide the pronunciation (e.g. of a newly-created word formed by combining a couple of kanji) – and of course it’s much better if words and their kanji match in a systematic way – kanji don’t strictly prescribe the pronunciation. It’s more accurate to say that the word itself does.
E.g. I recently learned the word for ‘speech’, 台詞, which – like other part-of-speech-type words that end in -詞 tend to end with the reading -し – I expected would probably be pronounced something like だいし or たいし.
But in this case, this word for ‘speech’ has a seemingly completely unrelated pronunciation, せりふ. And neither 台 nor 詞 are listed (in Jisho.org, at least) as having any readings in any way similar to any part of せりふ.
So, however it happened, historically, or whatever, the word is pronounced せりふ, and that’s that. The kanji, in this case, are more or less just ‘along for the ride’, though they do carry some semantic meaning.
So, even in this case, I wouldn’t consider this a case of the kanji having ‘exceptions’ or ‘inconsistent’ readings. There are, according to Jisho.org, at least (I haven’t done any research on this), no official or prescriptive (even if uncommon) readings of 台 or 詞 that give the kanji a ‘reading’ of せ, り, ふ, せり, りふ, or せりふ.
It’s the word that is has the ‘reading’ せりふ. And the fact that this word has this reading and this kanji, doesn’t automatically imply that now the kanji also have any such ‘readings’. In this case, the kanji are being used for some other purpose than ‘reading’. (Presumably semantic, but who knows (I don’t!), there could be a bazillion possible historical/accidental reasons how this word, its reading, and its official kanji came to be conjoined together.)
Maybe I’m wrong, or maybe there’s a more logical way to express the correct relationships, I don’t know. Maybe it truly is an example of an ‘exception’ in Japanese. But to me, it doesn’t feel that way. To me, the reading, expressed in Hiragana, is せりふ. That’s the reading given, and that’s also how it’s pronounced when spoken. The kanji ‘doesn’t get a say in the matter’, it seems to me, so it’s not really a difference between how the word is ‘read’ in terms of characters and how it’s ‘read’ in terms of pronunciation.
Anyway, hopefully that example helps to illuminate why people were disagreeing with what you wrote about kanji having differences/exceptions between textual and verbal expressions. (Again, I could easily be wrong; anyone feel free to correct me if so, please! )