Are you an intermediate Japanese learner struggling with reading kanji? In this article, I go over a simple but important sound change that occurs when kanji are used together in compounds, and a rule that can help predict how they should be read.
Studying Through Reading
I’ve been studying Japanese on and off ever since I became interested in Japan – for over 20 years now. After getting to an intermediate level, I reached a plateau. No matter how much I talked, I seemed to be recycling the same phrases, having the same conversations, and making the same mistakes. New words and expressions refused to stick. I felt disheartened and, as a result, started having fewer conversations.
A few years ago, I had a breakthrough enabling me to move forward again. Although I am still not as fluent as I would like, I once again started to feel small but definite improvements in my ability. The breakthrough was to stop worrying about speaking and focus on reading. I like reading books and comics, and one of my objectives for studying Japanese was too to be able to read manga in their original form. I decided to make that my main purpose.
For this, I use a two-pass method. First, I read a volume of manga as fast as I can, focusing on general comprehension only, but marking unfamiliar words with a pencil. Later, I do a second pass, looking up new words, and turning them into flashcards for Anki on my computer. I’ve been doing this for several years, and I’ve accumulated nearly 10000 flashcards, averaging a hundred revisions a day. In addition to improved reading ability, I started noticing patterns in the way that consonant sounds change when two or more kanji are combined into compounds.
Sound Changes Have a Name
Kanji usually have two different readings: a Chinese one (“on yomi”) and a Japanese one (“kun yomi”). As a general rule, standalone kanji use the Japanese reading, and those within a compound use the Chinese one, but not always. Sometimes compound kanji use Japanese readings or combine Chinese and Japanese ones. This is why reading Japanese can be a real headache. These sound changes mostly concern compounds that use Japanese readings.
Let’s look at an example: the kanji for “to read” 読 “yo.mu” (the 2nd part is the verb ending which can change). However, inside a compound, it’s read as “doku”. So “reading” (the activity) is “doku sho” (read+write). In a sentence it becomes:
So far so good. With time, effort and practice, this hurdle can be overcome and juggling the Japanese and Chinese pronunciations can be done in a satisfactory manner. However, things are never so simple. Occasionally, the first consonant sound of the second kanji in a compound would change. Let’s look at another example, this time a compound that uses Japanese readings since sound changes mostly occur with them. The kanji for hand (“te”) can be combined with the one for paper (“kami”) to form the word letter (“tegami”). Here is an example sentence:
In our compound, “kami” has become “gami”. In Japanese, this sound change is called “rendaku” and affects consonants such as K, S, T, and H. The change is equivalent to adding the “dakuten” ( ﾞ ) mark to Japanese syllables starting with those sounds, and turning them into the consonants G, Z, D, and F. I will refer to the former set as “unvoiced consonants”, and the latter as “voiced consonants”, although in reality, the distinction between the two sets is more complicated. There are more sound changes, but let’s stick to those to keep it simple. Let’s look at examples of each:
A Century (or Two) Old Law
My problem was that this change didn’t happen every time. No one ever told me about this at my Japanese school (the focus was more on speaking than reading). What are the rules for this? Is there even a rule? Do I need to memorise each sound change individually? At first, I did, but I kept track of the times the changes did and didn’t occur. I sensed that there was some pattern but I couldn’t quite put my finger on it. At last, I ran a Google search and found something called “Lyman’s Law” buried away in some Wikipedia article.
Benjamin Lyman was an American Mining engineer who was one of the many foreign advisors hired by the Meiji government in the 1870s to help modernise the young nation of Japan. He surveyed oil and coal deposits, and published the first geological map of Hokkaido in 1876. He was also an amateur linguist, and, like myself, he also noticed a pattern with the sound changes described above. In 1894, he formally stated the necessary conditions for a “rendaku” sound change to occur. These had previously been discovered by Edo linguists a century before.
Simply put, Lyman’s law states that a kanji cannot contain more than one “voiced consonant”. In other words, if the second kanji within a compound, starts with an “unvoiced consonant” but also contains a “voiced consonant” than there is no sound change. In a way, this makes sense: saying two “voiced consonants” in quick succession can be tricky so best to avoid it whenever possible. It’s time to look at a couple of examples:
Looking at the second kanji each time, in the first example, the “T” of “tera” changes to “D” since “R” is not a “voiced consonant”, and in the second one, the “K” of “kado” remains unchanged since “D” is a “voiced consonant”. It’s important to note that other conditions are necessary for a sound change to occur, and there is no full set of rules that can completely predict them. Lyman’s law is really just a guideline. One interesting observation is that the longer the kanji reading, the more likely there will be a “voiced consonant”, and the less likely a sound change will occur.
Getting Sound Changes Right
Hopefully this information will be helpful when reading Japanese, and can save you time when looking up words. Over time, you should acquire a feel for when a sound change occurs (or not), and you will also start to use them naturally when conversing. You might even be able to impress your Japanese teacher (and classmates) by getting them right most of the time.
Finally, it’s important to note that although you may feel that saying “teKami” instead of “teGami” is no big deal, surprisingly, many Japanese will find it hard to understand. This is perhaps because there are many similar-sounding words in Japanese, so trying to match a meaningless word to a similar-sounding meaningful one is much tougher in Japanese than it is in English. In a way, it’s similar to using “a” instead of “the” or omitting the article altogether, a common mistake by Japanese speaking in English. It’s not a big deal but does make me feel confused every time I encounter a misused article.
If you want to read more, Tofugu has an excellent and detailed article on this topic: Rendaku: why hito-bito isn’t hito-hito.