To some extent, learning Japanese is like learning at least six different languages.
It was originally a home-grown spoken language (though I have heard theories that it was influenced by Turkish–wrap your brain around that for a moment). Early on in Japanese history (as we understand “Japan” as a country, not just an occupied group of islands), it imported written characters called “kanji” (漢字) from Chinese, a very developed and literary society. These characters carry both pronunciation and basic meaning, and when you put them together with other characters, you can express more complex ideas. They fit the Chinese language beautifully, even though the sounds or even shapes associated with them have changed throughout the history of their use. They do not, however, fit Japanese quite as perfectly.
The spoken Japanese language adapted this written language by throwing the characters around, mashing them together, smashing them a few times, and then pulling the tangled mess apart. What they wound up with was a reasonably reliable method of recording written language, a large number of loan words from Chinese, and some fragments of kanji that wound up being devoid of meaning and would only represent phonetic sounds: hiragana (ひらがな) and katakana (カタカナ).
When Japan was primarily in contact with countries now considered China and countries now considered Korea, this linguistically served them fairly well. Then the Portuguese missionaries and Dutch traders came in with entirely foreign ideas and items that had no existing words to describe them. Furthermore, when Japan chose to start Westernizing, they sent students out to developed countries in the West, and they came back from English-speaking and German-speaking countries and then some with even more ideas to express. These words of Western origin were adopted into Japan by way of katakana syllabary.
As contact with foreign nations increased, so did the percentage of the Japanese language that was based on loanwords. However, like trying to get a written language like kanji to fit the spoken Japanese language, it took smashing and stretching these words a bit. For instance, my name, Brittany, is written as ブリタニー, and said Buritani (Boo-ree-tah-nee). Whether or not this is recognizable to people who don’t speak Japanese (or at least who don’t speak Katakana-English!) is anyone’s guess.
Hence, in order to speak Japanese with any fluency, you sometimes have to relearn English (and other languages) in a different way.
There are a few words that trip me up sometimes. For the record, “harmonica” is “haaah-moh-nee-kah”(ハーモニカ). Another word that trips me up because it’s so hard to recognize out of context is Connecticut (“koh-neh-tee-kaTTOH”, コネティカット). If we’re talking about US states or geography it’s easy, but when someone asks me in an open question-answer setting, “Do you know about Connecticut?” I risk looking like I know nothing about my home country’s geography when I have to ask what it is. Yes, I bring this up because it’s happened–twice.
Imagine how much sillier I look when I say, “Makudonarudo (マクドナルド)? No, I don’t think we have that in the US.”
Disclaimer: This is from the perspective of a native American English speaker whom has been actively speaking and studying Asian languages and cultures and histories for most of the past decade, especially those of Japan and China. I do not formally study linguistics, therefore this is not presented from the point of view of the study of linguistics. Yes, there is a difference.