Recently, a teacher of an alphabetic, phonetic language who does not teach Mandarin posted online to the effect that:

heavy intellectualizing teachers are driving the Mandarin train further down into the intellectual caverns of non-productivity.

This, at least, afforded me a good laugh. The teacher’s premise was that concentrating “too much” (in that teacher’s opinion: at all) on early reading of characters was supposedly sucking the joy from Chinese learners, and preventing them from experiencing the joy of just freely interacting in the oral language.
Maybe so. But anyone who has actually studied Mandarin to proficiency (even more so than those native speakers who acquired it and then later learned to read it — they did not experience the same issues) will know that the longer you delay reading, the harder it is, because it means significant memorization is required. Most of us had a commuter pass on that train to non-productivity in reading Chinese, back when we learned the language.
Memorization doesn’t sound like joy to me. It sounds like — intellectualization. It sounds like the reason so many students drop out of Mandarin. It’s definitely not the spoken language that’s the issue. It’s nearly always the reading and writing piece — literacy — which is not acquired but learned, no matter what language we’re talking about.
Personally, I see nothing “non-productive” about first-year students, starting with no Chinese knowledge, who can easily and fluently read 130 characters in any combination you care to put them, whether it be a story they’ve created together, a parallel story, or even a dialogue format that they have never seen before. They’re not decoding. They’re reading. They self-correct. They argue about missing characters. That is a facility with reading among these students that I have never observed or heard of through any other means of Chinese literacy training. If you have not contrasted these students with students who memorized characters to learn to read, you wouldn’t understand the degree of this difference. And this is not just my opinion (that they read easily, that they can read unknown texts, with no prior explanation of the content, that they self-correct) — this is what six years of data collection shows.
I think many people who are lobbying for “freedom” in CI teaching don’t know what Chinese teachers are doing when they teach early character reading through Cold Character Reading. I would be the first one to discourage “teaching” characters right away. Before I stumbled upon CCR, I used to delay reading for that reason — I didn’t want to make Chinese class all memorization. I believed there was no way to have kids read that would not be pure memorization and too onerous for the early days. But in CCR, in fact, we’re not  “teaching” characters at all. We give the student a “Chinese voice”, and the voice allows him to read characters. The problem is that, much like Comprehended Input itself, it’s just too simple for anyone to take it seriously until they really know what it can do.
The traditional way actually teaches characters. You might get a list of 30 or so words (which meant about 50 different characters, since many words have 2 characters in them) and you go home and memorize them and are quizzed on them. Very often you would be required to turn in a handwritten sheet on which you copied each of those characters anywhere from 25 to 100 times depending on your teacher’s degree of zeal. I remember (and am backed up by my diary from those days) calling home from Taiwan in tears while I was at language school there as a college junior (and in those days, a call to the US meant $40 at a time when dinner cost $0.50, plus a trip to the phone center, not just casually Skyping away at home) because I could not finish my homework and felt I was failing my family and everything they had invested in me to get me to that point. I couldn’t finish the mechanical copying work in order to get to the “make sentences” part of the homework — assuming that work would get me to read and write Chinese. My diary shows that I was seriously thinking about getting out of Chinese. I remembered the phone call, but until I read that diary, I didn’t remember that I was ready to throw it all in and quit Chinese — because of the way literacy was taught.
People who don’t teach Chinese aren’t aware of these things. They are also not aware that by and large these things have not changed significantly. And there are teachers out there, even teachers who embrace CI otherwise, who are still teaching literacy this way. There are teachers who have done huge amounts of TPRS training and who are widely regarded as fantastic story-askers, who just cannot let go of the character teaching thing. The memorization. The copying. The character copying is dressed up now, with pictures and cartoon characters maybe, and the flashcards are slicker, but students are still often told that the only way to become literate in Chinese is Hard Work. Memorize. Memorize. Copy. Memorize. Flash cards are electronic now, instead of being easily spillable piles of index cards on your desk, or a lump with a rubber band in your back pocket. I went for years never going anywhere without a large clump of character flashcards on my person. Literally for years. And I know of no student of Chinese — apart from the small number taught with TPRS and CCR — who has not had this experience.
Cold Character Reading represents a radical departure from this legacy model. It rests entirely on comprehensible input (so saying it goes against CI is a bit of a leap). It relies on giving the student a “Chinese voice” that will whisper in his ear while he’s looking at these characters. Those characters give him absolutely no information about the sound of the words they represent. That “voice” lets him know what the words are. CCR provides students with lots of opportunities to let that voice guide them in reading, until their eyes are familiar with the characters that represent the language they have acquired for comprehension and they are truly reading — not decoding. The large amount of reading means that we have taken all those flash cards and all the myriad times flipping through each stack, and replaced it with seeing those same characters in unpredictable contexts, over and over. Doing it in context rather than on flashcards means there are language gains at the same time. It’s a win-win.
The “price” for this — if indeed it is a price, because acquisition hardly seems a bad thing — is giving students dense comprehended input so they acquire the language they will read. By “dense”, I mean repeated. Deliberately concentrated and repeated input, usually accomplished through the frequently vilified technique of circling. And by “acquire” for the purpose of reading Chinese, I mean “acquire to the point where the language is very easily comprehended”. That requires a pretty high degree of repetition for beginners in Chinese, because the most frequent comment by learners is “all the words sound alike, and most of them start with /ch/.” This repetition of comprehended input in unpredictable ways is what gives these students the “Chinese voice” that whispers in their ear while they are reading, telling them what the next word must be or warning them that it cannot be what they have just said, prompting them to think about the meaning and get the right word.
But aren’t we killing at least some of the joy by only “allowing” students to use certain words in their class discussions and stories? I suppose we might be — if that were the case. But I’ve never heard of anyone Ci-based training Chinese teachers never to use any word that is not in a reading. The oral input piece for Chinese is optimally classic TPRS, which has all sorts of provisions for students to shape the language that is used in class. You can use any language you like as long as it’s made comprehensible and comprehended at the moment it’s used. But if that language is to be read, it must be acquired for the purpose of comprehension (not necessarily full production) — and that requires repetition.
This is where Chinese is different from Spanish. This is where Chinese reading is different from Spanish reading. This is what the Spanish teachers don’t “get” about Chinese, because it’s not a problem they need to deal with. (Just as, to be fair, most Chinese teachers don’t “get” conjugation issues!)
When you read Spanish, if you can recognize the word when it’s spoken, you can read it. You see “perro” [dog] and you know what that sounds like. You don’t need to be able to produce the word “perro” from zero on your own, because the alphabetic script gives you that lift. At the very worst, you might confuse it with “pero” [but], but context would sort that out pretty fast.
Chinese isn’t alphabetic. It’s not phonetic, either, for anyone but a skilled reader with a huge oral vocabulary. The layperson believes the hype out there about how Chinese radicals and phonetics make up such and such a high percentage of characters in the language (that part is true) and so (here’s the part that’s not) learners of Chinese can use them to figure out words. It doesn’t work that way. As students become more advanced — by which I mean ACTFL intermediate-high and above, recognizing perhaps 2000-3000 characters) these character components become more useful in reading, but there is still no way to know which word is represented by a character or two characters (many words have two, remember) unless one already knows the word orally and can say something like “Hmmm….that’s a food radical, so this word has something to do with food…and that phonetic part is going to be kang…or gang….or perhaps hang [sorry, folks, but phonetics aren’t a one-to-one correspondence in the first place) so what word do I know that has something to do with food and sounds like that?” It’s unlikely that someone learning Chinese who has a fairly basic word stock is going to be able to luck into recognizing a radical-phonetic combination that rings the bell for a word he already knows.  We still pop these elements up as useful things in recognizing characters that represent known words, but at the TPRS level, we don’t expect students to be able to read words that have not been acquired to the point of instant comprehension.
To read Chinese, then, there are two choices. One, memorize every single character and combination you’re going to see, and then you can recognize them. That’s not joyful at all. That’s “hard work” and that excludes those who are not good at school. That drives students who aren’t “good at learning” out of language, which is not the message we send with CI instruction. But you could conceivably (it has been done frequently in the past) delay reading for a year or even more, and then just have students memorize the characters for the words they then know well. And the minute you start doing that, you start losing students. There were 60-some students in first-year Chinese when I started in university, and seven of us graduated still taking Chinese. And our university didn’t wait to read, so the memorization was just steady and “average”, not crippling.
The other choice is CCR. Use concentrated, dense, repeated comprehended input to produce a “Chinese voice” in the student’s head, and let that “voice” help the student read directly in texts. Texts that have a manageable number of new glyphs in them. Not texts with hundreds of characters the student hasn’t seen before.
I am not saying, and I have never said, that CI doesn’t work for Chinese. It works just dandy for people to acquire spoken Chinese, as it does for them to acquire any spoken or signed language. What I am saying is that literacy is not language. Language is acquired. Literacy is learned. That’s the most basic distinction in the CI world, and anyone who doesn’t get that needs to go back and get it. Learning Spanish literacy isn’t hard as students already know most of the sound-letter correspondences. Learning Chinese literacy is a little more complicated and requires a different approach, because Chinese is different due to its writing system.
What I am saying is that memorization SUCKS. We who teach CCR do not want to kill out students’ joy in literacy, which — unlike language — must be learned to a greater or lesser degree depending on the opaqueness of the script. The Japanese teachers I am working with feel the same way, and are experiencing similar results.
So if one can abandon an insistence on “total freedom of input”, which is based on a false dichotomy of “free is joyful, any restriction on input kills joy”, one might understand that teaching characters early through CCR is the joyful, “non-intellectual” option for Chinese.
At the end of the day it comes down to some simple facts. Mandarin teachers have to choose and in so doing, they will have to potentially “kill joy” in one place or another (if that even happens, which there is no proof of in the first place).
1. They can give the supposed joy of total freedom in input — all free, don’t worry about repeating, etc.
Then they will need to crack the whip when the students have to memorize 100+ characters to read anything after they delay reading for awhile so as not to “kill the joy” of that oral input. (100 unique characters will be maybe a 60 word level story. Remember, there is often more than one character to a word. And it takes a lot longer to get kids to a 60 word level in Chinese than it does in Spanish, where you can do it on cognates alone.) And realize that the reading you get this way will be decoding because there won’t yet be a Chinese voice to make it make sense, because that total freedom on oral input meant no dense repetition. Or, alternatively, one could add another 200 or so characters to the instantaneous memorization load and wait until the end of a year for diffuse input to create a voice. But with so many unknown characters, the “voice” will lose the power it has when there are relatively few unknown characters to be dealt with. Either way, you are left with memorization of characters and  weak or nonexistent “voice” which means decoding, not reading.
2. Teachers can “deny” the joy of total freedom during oral input. They can use classic TPRS with personalization. Again, it is a false dichotomy to say “freedom=joy, any restriction=no joy”. Classic TPRS provides many opportunities for students to hear and use vocabulary of their own choosing, and I know of no evidence proving students don’t feel ownership in these co-created stories simply because the teacher deliberately included a particular high-frequency word in them and repeated it skillfully and unpredictably. So students still have a significant dose of joy during oral input, and then we can give our students joy during reading as well because we eliminate the memorization, the write-it-50-times, and so on. They have the joy of achievement, the joy of extensive reading of stories they like (we are not teaching them to read with Quotations from Chairman Mao or oracle bone inscriptions, after all). And they get the benefit of additional language input as they are learning to recognize characters — unlike memorization-based techniques during characters are learned separately, word-by-word, with little or no context.
So, understanding the nature of reading in Chinese and how it has been done to foreigners for thousands of years, it’s really hard for me to see how anyone could call cold character reading “intellectualization”. It’s precisely the opposite. It’s the first approach that has said “These squiggles are language. Let’s give you some language and let your brain practice pattern recognition in a no-stress way while still getting reinforcement on all kinds of language features.” That sounds like great CI to me.