Rendaku (literally “sequential voicing”) is one of the two core concepts of the formation of jyukugo, the other being choosing the proper readings. In essence, rendaku is the alteration of a kanji reading, under specific circumstances, to turn an unvoiced consonant into a voiced consonant. This is done because, and solely because, this makes the word easier to pronounce.
For example, the word 出口. This is pronounced “でぐち”. or “deguchi”. But in actuality, the reading for 口 is “kuchi”. So what the heck is going on here?
If you’ve been studying Japanese for any length of time, you’ll have noticed that jyukugo, or compound words, mostly follow this pattern, but it may not be clear what’s actually going on. What’s going in is that the vowel in the previous mora, “e”, blends better with “g” than “k”, so they change to a voiced vowel to make the transition easier. “kuchi” is the same word as “guchi” in this case, the vowel has just been transitioned to voiced for ease of pronunciation.
This is very common and mostly predictable, but there are situations where it does not occur. See the Wikipedia article for further information. Again, we’re not intending to be a compendium of obscure information, but rather to make you aware of concepts that you may not be aware of or familiar with. See our page on Handaku for other examples of how consonant voicing can change.