Most would agree that the most difficult thing about Japanese is the kanji. And sure there’s over 2000 of them – over 3000 if you want to master all the kanji used mainly in personal names and other uncommon uses. But for general literacy, just over 2000, and it doesn’t really help a lot to know most of them; you really need to know all of them.
So the temptation is to think, “Well, kanji is pretty advanced; I’ll learn what I can in spoken communication, and use kana for reading (or – shudder – romaji).”
Do not do this. Do not even think this.
In fact, for someone just setting out to learn Japanese, the kanji should be the very first thing to master, perhaps even before you learn your first vocabulary word*.
The reason for this is that kanji make reading a lot easier. Easier? really? yes. In fact, a sentence written entirely in kana can be nearly incomprehensible. Where does one word end and another begin? Well, if you have kanji, it’s immediately obvious. But more than that; what does this particular word mean, when there’s a dozen homonyms for this string of three kana?
If you think i’m exaggerating, consider the case of the word(s) こうか. Take a wild guess how many different words (not shades of meaning) are represented by this brief string of characters. Five? Ten? surely not more than a dozen, no? Wrong – there’s twenty-three**. Twenty-three! Now granted, the intonation will be different between many of them (but you can’t tell that when reading); and granted, the context will be a massive help; and granted, this is an extreme example. But what if you know the kanji? Then this case becomes trivially simple, because each of these words is written with different characters.
Learning vocabulary is far simpler when you are already familiar with the kanji. Since you have the meaning attached to the visual code of the character to hang your memory on, sometimes you only need a very few reviews to remember a word for life. Often enough you can make a very good guess at the meaning of the word the first time you see it. Without them though, you’re swimming in homonyms and characters that make no more sense than hieroglyphics.
I read somewhere that the Japanese (and, I would presume, the Chinese) use an entirely different part of the brain for language than other language groups that use alphabets or syllabaries. I could easily believe this, because relating those visual symbols directly to meaning without the intermediate phonetic step feels very different than scanning a string of letters that, other than the order, all look the same.
(Current status: at present I’m at 950 cards on the RTK site, so almost halfway through the first step, which I expect to finish sometime in June.)
* you will note that I am, unsurprisingly, a hypocrite. In my defense, I came to this conclusion rather later than I would have liked to, after a considerable degree of frustration at mixing up words, not remembering words I ought easily to have recalled, and not being able to read words I “knew”. As soon as I realized what was going on (yeah, I’m a little slow) I started to devote nearly all my dedicated study time to kanji.
** taken from results listed in Kotoba!, the dictionary I mainly use. Highly recommended, by the way.
Lan'dorien's Mysterious Journey 
A very interesting read.
The reasons that you described above are exactly the reason why spoken Japanese is sometimes so difficult to master for non-native speakers. As context is massively important in Japanese, without understanding the rest of the sentence surrounding a word such as こうか then you are totally at a loss as to what is being said. Consider the same in English, where grammatically structures are less rigorous, yet less of a sentence need be understood in order to comprehend the entire meaning.
This is further compounded by the fact that the Japanese tend to drop as much of a sentence as they possibly can and rely solely on context-dependency to get across their meaning. So you can have some sentences that are boiled down into just a couple of words, as everything before that particular sentences refers to the sentence being said. Again, this is not helpful in a language where homonyms are firmly in control in the first place!
Your point about the part of the brain that is used by Japanese speakers is also an interesting one. When using a sign language to communicate, the part of the brain that stores and activates the language is completely different to that which is used by users of spoken languages. There was the case in the UK a few years back of an interpreter of British Sign Language that had a stroke, and although he could no longer speak, he was still able to converse via BSL as that part of his brain was left unaffected. Therefore, I think in times past, when kanji was definitely more representative of the pictures from which they were derived, then yes, perhaps a different part of the brain was used, but I wouldn’t say that was as much the case now that kanji have all but lost their original forms (though I could be completely wrong, obviously!)
Again, a really interesting read!
Thanks!
That development of kanji you mention, from simple pictographs to highly abstract symbols, really fascinates me. It reminds me of the part of Gulliver’s Travels where Gulliver encounters a group of scholars who attempt to communicate without using words at all. Instead, they carry enormous packs containing every object they wish to talk about, and lay the objects out in order – it’s all very well until you need a verb, or worse, an adjective!