7

I read once (in this comment by Victor Mair on Language Log) that Chinese has single morphemes that span two hanzi. The example given was the Chinese word pútáo 葡萄. At the time, I assumed it applied to Japanese equally, because I assumed 葡萄{ぶどう} was the same word. I then assumed I could generalize from that to similar compounds. (In retrospect, I don't think that reasoning was very good, which is why I'm asking this question.)

Other compounds that look like they might be monomorphemic include [薔薇]{ばら}, [蜘蛛]{くも}, and 麒麟{きりん}.

Are any of these single morphemes? Pairs of bound morphemes? If these are bad examples, are there two-kanji compounds that are single morphemes?

  • 2
    That would depend on how you define "morpheme", which is not trivial. You should definitely be able to argue that all of your examples are single-morpheme, since for instance, it would be hard to argue the ば of ばら has any meaning by itself relating to "rose". – dainichi Oct 26 '12 at 04:28

2 Answers2

2

The only place you would find morphemes that span multiple kanji are in gikun or in reformed words, simply because "morpheme" in Japanese is defined as the sound a single kanji or kana has/makes.

「[海鷂魚]{えい}」

「[今]{け}[日]{ふ}」 -> 「[今日]{きょう}」

Community
  • 1
Ignacio Vazquez-Abrams
  • 4,989
  • 2
  • 22
  • 29
  • 2
    Also 百舌鳥, the first one of these types that I ever saw – ssb Oct 25 '12 at 23:49
  • 1
    To add to Ignacio's clear and concise answer: gikun are cases where the reading does not match particular kanji in the compound... and sometimes has more kanji than morphemes, implying that at least one morpheme would cover two kanji (although the common view is that there is simply no kanji<->morpheme connection for such compounds). I am less sold on 'reformed' words: even the example above doesn't really show two kanji for one morpheme (merely a blurry frontier). – Dave Oct 26 '12 at 01:24
  • 7
    "morpheme in Japanese is defined as the sound a single kanji or kana has/makes"? This is utter nonsense. So the morphemes of a word depends on whether you write it with kana or kanji? – dainichi Oct 26 '12 at 01:41
  • ssb: You must living in Osaka, like I did, 百舌 (more specifically 中百舌鳥） was my first too! – paullb Oct 26 '12 at 09:08
  • @Dave Morphemes are units of meaning, not sound. Counting morphemes has nothing to do with counting kana sounds. – kandyman Dec 13 '17 at 15:25
  • 1
    @kandyman, I'd learned _[morpheme](https://en.wiktionary.org/wiki/morpheme)_ to mean "the smallest unit of meaning", as pertains to the **spoken** language. With Japanese in particular, there can be yawning gaps between the spoken and written languages. The smallest unit as pertains to the **written** language, would be the _[grapheme](https://en.wikipedia.org/wiki/Grapheme)_. See Wikipedia [here](https://en.wikipedia.org/wiki/Grapheme#Types_of_graphemes), particularly the sections _Types of graphemes_ and _Relationship between graphemes and phonemes_. – Eiríkr Útlendi Mar 16 '18 at 20:53
  • FWIW, in Japanese, there _is_ some overlap between the units of sound represented by kana, and morphemes. Consider こ -- this clearly expresses a morphemic property in words like これ、この、こちら, in contrast to the semantic value of そ in words like それ、その、そちら. This overlap is not consistent nor always one-to-one: this こ could have very different meanings, as in あきこ、コンピュータ、鎖国【さこく】. – Eiríkr Útlendi Mar 16 '18 at 20:57
  • There is often overlap between the units of meaning represented by kanji (**graphemes**), and the corresponding meaningful units of sound (the readings of those graphemes). 百円 = "_hyaku_" + "_en_" reading = "hundred" + "yen" meaning. In other terms, the graphical representation has nothing to do with the spoken representation: 百舌鳥 = "hundred" + "tongues" + "bird" for the meaning, but "_mozu_" for the reading, likely from Old Japanese _momo_ "hundred" + _su_ "bird". The spelling and the reading have independent derivations, and modern _mozu_ is a single unit and semantically inseparable. – Eiríkr Útlendi Mar 16 '18 at 21:06
  • @EiríkrÚtlendi I'm afraid you are confusing the matter when you mention spoken language vs written language. Spoken language is merely a vocal manifestation of the language, as is writing. The language itself consists of the morphemes, the lexicon, the syntax and the grammar. Whether you write it or speak it is irrelevant - the language has underlying properties which can be analzyed. The underlying morphemic content of Japanese is not affected by whether you speak or write a word. You mention こ as an example - its morphemic content entirely depends on whether it is 此、国 、こ, etc. – kandyman Mar 18 '18 at 20:33
  • @EiríkrÚtlendi Graphemes are the vehicle by which a language can be graphically represented. Long before Japanese had a writing system, it can a fully functional language - it had zero graphemese but still had a complete system of morphemes, syntax, lexemes and grammar. Graphemes are a way of expressing language, they are not the substance of the underlying language. – kandyman Mar 18 '18 at 20:39
  • @kandyman, in Japanese, a "unit of meaning" is one thing spoken, and another thing written. A kanji is a unit of meaning, but it has no immediate connection to morphemes. Is 承 equal to _shō_ (which is not a meaningful unit of sound in Japanese in isolation, and thus not a morpheme), or is 承 equal to _uketamawari_ (which is meaningful, but which also consists of [arguably] three morphemes)? The written / spoken distinction is very important when talking about "units of meaning" in Japanese. – Eiríkr Útlendi Mar 19 '18 at 23:51
  • @EiríkrÚtlendi I'm sorry but I think you are clearly misunderstanding what morphology is. Saying that "kanji...has no immediate connection to morphemes" is simply incorrect. You seem to be equating phonemes with morphemes. 承 is a morpheme, while shō is a phoneme. Each kanji is a morpheme. A word which is composed of two kanji is a word composed of two morphemes (with exceptions noted in this thread). The same for a word with three kanji. Thus, if you were familiar with the individual morphemes in a word, you could probably guess the meaning, like (安楽死). – kandyman Mar 20 '18 at 14:25
  • @EiríkrÚtlendi Another way to look at it is to use English examples first. Take 'un'. You can make the sound 'un' with your voice, but depending on what comes before or after that sound, it may or may not be a morpheme. If you say 'unbelieveable' the 'un' is the morpheme meaning 'not'. If you say 'sun'. the 'un' is not a morpheme, but is merely a sound which makes up another morpheme. If you say "shō" it depends on whether it is part of the word 承知 or 商売 or something else. It is not the actual sound which has meaning - the meaning is derived from what the sound is mapped onto. – kandyman Mar 20 '18 at 14:52
  • @kandyman, when you say that 承 is a morpheme, suffice it to say that you and I have divergent views. Comparing Japanese writing to English writing doesn't really work, as English writing is broadly phonetic (granted, some multi-graph spellings are quite odd, but on the whole, the glyphs and glyph combinations are directly connected to specific sounds), while Japanese kanji are much more abstract (in that a single kanji can have umpteen different readings). In the term 百舌鳥, we have three graphemes, but only one morpheme _mozu_, which is largely disconnected and independent of the three kanji. – Eiríkr Útlendi Mar 20 '18 at 16:26
  • @EiríkrÚtlendi You have given an example of an exception (百舌鳥), not the rule, as noted above. If you do not believe that the kanji 承 is a morpheme, I'm afraid to say your view does not match standard linguistic theory. Consider how kanji came into the Japanese language. Characters which represent specific meanings were mapped onto Japanese words which represented the same or similar meanings. Therefore kanji are fundamentally connected to the meaning of both the native Japanese words and the imported Sino-Japanese words. They are the basic units which deliver the meaning of Japanese words. – kandyman Mar 20 '18 at 22:42
  • @EiríkrÚtlendi Perhaps this is not the right place to have a detailed discussion. If you wish to discuss the issue more, I am open to alternative methods of communication. – kandyman Mar 20 '18 at 22:43
1

Chinese is a lot neater with regard to its characters; one character equals one word (now morpheme) equals one syllable. In theory at least. 葡萄 being a two syllable morpheme, Chinese would rather adhere to a policy of one character per syllable than one per morpheme if it has to choose. In ancient Chinese there were prefixes and suffixes as well, but they could be added without changing the number of syllables, so one character per syllable held true. All of your examples look like borrowed words where this happened. Japanese is a bit messier. Niwatori is evidently 庭＋鳥、but since in Chinese "chicken" is just one word, Japanese follow suite and uses 鶏. I know there are cases of the opposite, where the word has more kanji than morphemes, or even syllables sometimes, but my Japanese isn't great, so I wouldn't be able to give any examples. Someone mentioned 百舌鳥.

Earthliŋ
  • 47,707
  • 9
  • 125
  • 198
lin
  • 26
  • 2