Chinese Character Acquisition: A Native Perspective

In Uncategorized by Skritter

author photoThis week’s blog post is in response to a question submitted by one of our Skritter members, Drew. He wrote us and asked:

“I am always fascinated by how native Chinese speakers identify characters by referencing the pronunciation and then radicals, but I never really understand how to do it myself. Would you be willing to do a blog post on this?”
The Question 

So how do they do it, and can we as non-native speakers do it too? I asked a few locals here in Taiwan, along with two of my former Chinese teachers, about their character identification tricks, and here’s what they generally said:

I just look at the radical and the phonetic component! Most modern Chinese characters are phonograms: one part tells the meaning of the character and other part identifies the pronunciation. (translated and paraphrased into English)

Well there you have it Drew, native speakers just look at the radical and look at the phonetic, and they’re done–SO SIMPLE! 

The Facts

Yeah right. If only it was as easy as 1+1=2. Figuring that Drew wouldn’t be happy with such a response, I decided to dig a little deeper into the question. I checked out a research paper titled: “The Acquisition of Chinese Characters: Corpus Analyses and Connectionist Simulations,” which was published by Beijing Normal University in the Journal of Cognitive Science in 2004. And here is what I found, condensed into five major points.

  • There are 5,631 phonogram characters accounting for 81% of the total 7,000 frequent characters (based on the National Language Commission of China and work by Li & Kang, 1993). 
  • While the phonetic of a phonogram does not (always) provide reliable pronunciation, it will relate to character pronunciation in three ways:
    1. Regular: the whole character is pronounced just like the phonetic.
    2. Semi-regular: the whole character is pronounced partly as the phonetic, with a different tone, different onset, or a different final. 
    3. Irregular: the whole character is pronounced completely differently from the phonetic. 
  • The consistency effect plays a large role in phenomenon of processing characters. This referes to “the degree of consistency in the pronunciation of the group of characters that share the same phonetic component.” There are essentially two possibilites:
    1. Consistent: all characters that share a phonetic compound are pronounced the same as the phonetic.
    2. Inconsistent: characters share a phonetic but are not pronounced the same. Some might follow the pronunciation, while others don’t at all. 
  • People (native speakers) only use a character’s phonetic clues on low-frequency characters. For high-frequency characters they are recognized on a whole-character basis. 
  • Elementary school textbooks contain 2,570 characters, and 74% of these characters are phonograms. 
The Conclusion
Phonetic components certainly help native speakers identify words that they don’t see often, but only when they are low-frequency (perhaps words that appear in classical texts, poems, etc.). Most words, however, are actually memorized over a long period of time. As this research indicates, it takes six years of schooling for native speakers to learn 2,570 characters, for example. Given the the general irregularity of phonograms (only 26% of modern Chinese characters are actually pronounced the same as the phonetic component), however, it is a skill that has a lot of “exceptions” to the rule. Therefore, native speakers might consider the phonetic, but they will also take into consideration other characters that share the phonetic component (see the consistency effect above), and doing so doesn’t mean they will always be correct.
I think that we as non-native speakers can learn something from this as well. Basically, the more characters that we acquire, the better we will be able to correctly identify the pronunciation, even when they are semi-regular phonetic or irregular phonetic pronunciations, because we are able to make connections to other characters that might fall into similar character groups. To improve upon our own awareness of phonogram characters we should be try to group characters based on phonetic components, making note of character groupings and inconsistencies. This isn’t something that is going to happen overnight, but once we have a solid foundation, we too should be able to pronounce new characters correctly without looking them up… at least some of the time, anyway. 

Have questions, comments, feedback, or a topic you want covered on the Skritter blog? Please send them to
Talk about this post on our forum!