We use the hiragana syllabary to write all the particles (analogous to prepositions like ‘to’, ‘from’, ‘with’, etc.), the suffixes of verbs and adjectives, and to write many words of frequent use, like greetings, even if they have a kanji, e.g., ‘chotto’ (one bit) has the simple kanji 一寸 but, still, its use is so common that we most often would find it in hiragana as ちょっと. We are also likely to write in hiragana words that have difficult kanjis, e.g., why write ‘tokage’ (lizard) as 蜥蜴, which takes 4 minutes, if we can write it in hiragana as とかげ, which takes 4 seconds? Thus, in general, we will only see unusual or many-stroke kanjis not written in hiraganas in places that need to save writing space, personal names, technical or academic publications, certain novels, etc. Reading kanjis is taxing even for Japanese people.

Learning the kanas is not difficult; Japanese children learn to read and write both hiragana and katakana in 1st grade, with time to spare to learn 80 kanjis. Learning the kanas is probably the biggest bang for the buck when learning Japanese; we can learn them with 1/100th of the effort needed to learn a small number of kanjis, and they immediately give us the possibility of writing anything in Japanese, and reading (i.e., spelling out) about 50% of it. Knowing kanas allow us to ‘read’ many mangas, but this does not mean that we’ll ‘understand’ them since, after all, they are in Japanese. Still, learning the kanas is a basic and huge step towards learning Japanese, and its really time well spent.

from kanji to hiragana

Hiragana started around the 5th century, at a time when kanjis were the only writing system available; kanjis represent meaningful ideas – things, feelings, actions, but since there are thousands of such meaningful ideas, there are thousands of kanjis. However, in the Japanese language there are only 110 different sounds, e.g., ‘a’, ‘ka’, ‘kya’, etc. Hiragana takes advantage of this small number of sounds: instead of using characters to represent meanings, it uses characters to represent sounds.

For example, to design the hiragana character of the sound ‘me’ we would select one of the many kanjis whose kun-yomi (Japanese reading) or on-yomi (Chinese reading) sounds ‘me’; one such kanji is 女 (woman, female), which has two kun-yomis: ‘onna’ (like in 女, onna – woman), and ‘me’ (like in 女神, me-gami – female deity), so we could use the kanji 女 to represent the sound ‘me’. In this fashion, we select 46 different kanjis to represent the 46 base sounds of Japanese; the remaining Japanese sounds are variations of the base ones.

Next we rewrite the 46 kanjis selected to represent the base sounds using cursive script. Similar to western script, these ‘scripted’ characters duplicate the entire original character making it more curvilinear and minimalistic, with fewer and simpler strokes; this makes it more difficult to recongnize the scripted kanjis as the non-scripted ones. For example, the cursive script of 女 is め. Now we set め as the hiragana character that represents the sound of the original kanji 女, e.g., め represents ‘me’. Hence, the resulting few-stroke fast-to-write hiragana characters often have lost the resemblance to the kanji characters from which they came from.

The following is the list of kanjis used to derive the base 46 hiragana characters, together with the resulting hiragana character [WP]. There is a clear resemblance between some kanjis and their hiragana characters, e.g., 仁 and に, 寸 and す, or 世 and せ; the resemblance between others is not lost, but it is more difficult to see, e.g., 計 and け, 由 and ゆ, or 太 and た; however, this resemblance is lost in many hiragana characters, e.g., 機 and き, 遠 and を, or 美 and み.

base characters

Hiragana has 46 base characters that give us the basic sounds. We can follow some of these characters with a small や (ya’), ゆ (yu), or よ (yo), to obtain 21 new sounds; for example, the character ひ (hi) followed with a small や (ya) becomes ひゃ (hya). This gives us a total of 67 sounds.

• Japanese has the sounds し (shi), ち (chi), and つ (tsu), instead of ‘si’, ‘ti’ and ‘tu’, which it doesn’t have
• は, へ, and を are pronounced ‘wa’, ‘e’, and ‘o’, when used as particles (see pronunciation)
• ふ is usually written as the sound ‘fu’, but its actually closer to ‘hu’, or to the English word who spoken just blowing air, without changing the shape of the mouth (see pronunciation)
• A character followed with the small ‘ya’, ‘yu’ or ‘yo’ becomes part of the syllable, which is pronounced in one ‘beat’, e.g., ひゃく (hyaku – hundred) is pronounced ‘hya-ku’, not ‘hy-a-ku’
• A continuous stroke in さ, き, and り is split in certain fonts and often in handwriting:
  さきり = さきり

ten-ten (“) and maru (°)

We superscript some characters with double quotes (“) or with a circle (°) to represent 45 more sounds. The double quotes are called ‘da-kuten’ (voiced-mark) or ‘ten-ten’ (mark-mark, or double mark), and the circle is called ‘han-daku-ten’ (half-voiced-mark) or ‘maru’ (circle):

• We can write ‘ji’ as either じ or ぢ
  in practice, we only use ぢ in a few cases like ちぢむ (to shrink) and はなぢ (nosebleed)
• We can write ‘zu’ as either ず or つ”
  in practice, we only use つ” in a few cases like つづく (to continue) and きづく (to notice)
• ‘ji + ya’ (じゃ) is ‘ja’ (like in ‘jazz’), not ‘jya’
• ‘ji + yu’ (じゅ) is ‘ju’ (like in ‘juice’), not ‘jyu’
• ‘ji + yo’ (じょ) is ‘jo’ (like in ‘John’), not ‘jyo’

Having both じ and ぢ sounding ‘ji’, and ず and つ” sounding ‘zu’, is an arbitrary complication, but many languages have the same issue; in English, for example, ‘sent’, ‘cent’ and ‘scent’ sound the same, so we just have to memorize the correct spelling for each word.

The hiraganas have a total of 67 + 45 = 112 sounds; of these, two sounds – ‘ji’ and ‘zu’ – have two representations, so the original Japanese language has a total of 112-2 = 110 different sounds. This is small number compared to that of English and many other languages.

consonant voicing

The role of the ten-ten (“) is to voice voiceless consonants:

and

The role of the maru (°) is similar to that of the ten-ten (“), except that ‘p’ is classified as a semi-voiced consontant:

transition into voiced consonants

In a compound word, it is often the case that if the second word begins with a voiceless consonant, we turn this consonant into a voiced one. For example, consider the two words 田 (ta), which means ‘rice paddy’, and 中 (naka), which means ‘middle’; by themselves, we pronounce them as ‘ta’ and ‘naka’, but we could combine them into a single word as ‘ta + naka’ or ‘naka + ta’. The ‘na’ beginning sound of ‘naka’ is easy to transition into, so we always have that

\begin{align}
\mbox{ta}+{\color{red}{\mbox{na}}}\mbox{ka} &= \mbox{ta}{\color{red}{\mbox{na}}}\mbox{ka}\\
\mbox{た}+{\color{red}{\mbox{な}}}\mbox{か}&=\mbox{た}{\color{red}{\mbox{な}}}\mbox{か}\\
\mbox{田}+\mbox{中}&=\mbox{田中}\\
\end{align}

‘tanaka’ – in hiragana たなか and in kanji 田中 – is the most common Japanese last name.

As for the other combination, i.e., ‘naka’ + ‘ta’, the ‘ta’ voiceless sound is not so easy to transition into, so ‘naka + ta’ might be ‘nakata’, or we might voice the ‘ta’ into a ‘da’ and end up with ‘nakada’:

Both ‘nakata’ and ‘nakada’ are common last names, and we write them both as 中田, so a person whose last name is 中田 might pronounce it as either ‘nakata’ or ‘nakada’.

In general, a transition into a word that begins with an voiceless consonant might or might not require us to voice that beginning consonant. Sometimes both options are available, as in ‘nakata’ and ‘nakada’, but often they are not; convention and/or usage has decided on one version or another. Here are some examples when the consonant has and hasn’t been voiced:

practice words

And that is all there is to hiragana. Now some practice: