In linguistics, stem (and sometimes root, or base) is the word or part of the word (morpheme) you get when you remove all its suffixes, prefixes, etc., e.g. if you remove ~ing from "dying," you get "die," so "die" is the stem of "dying." What the term means exactly seems to vary from author to author, and these three similar-sounding terms—stem, root, and base—may mean slightly different things when used by the same author. Usually it's used to say that you get the stem by removing suffixes, then you add new suffixes to the stem. In Japanese, some notable stem-related terms are:
- The term "masu-stem" refers to the ren'youkei 連用形, which is what you get when you remove the ~masu ～ます suffix from a word, e.g.: tabe is the masu-stem of taberu 食べる, "to eat," because it's tabe-masu 食べます without ~masu. You can add ~tai ～たい, for example, to this stem: tabe-tai 食べたい, "[I] want to eat."
- The stem of i-adjectives is an i-adjective without its ~i ～い suffix, e.g.: sugo~ すご～ from sugoi すごい, "incredible." You can add ~sugiru ～すぎる to this stem: sugo-sugiru すごすぎる, "too incredible."
- The term consonant-stem verb refers to a godan 五段 verb, because such verbs' stems always end in a consonant. Also called "consonant verb," or "-u verb" because you remove the "-u" vowel to get the vowel-less consonant stem.
- The term vowel-stem verb refers to an ichidan 一段 verb, because such verbs' stems always end in a vowel. Also called "vowel verb," or "-ru verb" because you remove the "-ru" syllable to get the vowel-ending stem.
Definition
For reference, a technical definition of the terms:(Haspelmath, 2002:266, 274, glossary)
stem: the base of an inflected word-form.
root: a base that cannot be analysed further - i.e. a base that consists of a single morpheme.
base: the base of a morphologically complex word is the element to which a morphological operation applies.
In other words, the difference between stem, root, and base, is, first, that the term stem is only used when dealing with inflection (conjugation, declension), while base is used in any case, which means all stems are bases, but not all bases are stems; and, second, that a root consists of a single morpheme, while there's no restriction applied to stems and bases, this means that if a word is composed of three morphemes, for example, its stem could have two morphemes, which itself has a stem that's only one morpheme, and that would be the root of the word, i.e. a stem can contain a nested stem, but a root can't contain a nested root.
By the way, in a plant or tree, the "root" is the part underground that serves as the base of the whole plant, while the "stem" is the part above-ground where the branches (shoots) and leaves and flowers come from.
In Japanese, gokan 語幹, literally "word-root," refers to the part of a word that stays the same when it's conjugated.(デジタル大辞泉:語幹) Japanese mostly uses agglutination to conjugate words, throwing suffixes on top suffixes on word-roots, so the word-root itself rarely changes. The suffixes are called katsuyou-gobi 活用語尾, "inflectional word-tail (as in word-ending)," e.g. in sugoi すごい, sugo~ すご～ is the gokan, while ~i ～い is the gobi.
Note that here we have a bit of a mismatch in definitions. The definition of stem, base, or root we saw is about being the basic morpheme upon which some morphological operation is applied. If you get "dying" by applying "~ing" to "die," then it follows "die" is the basic stem of "dying." On the other hand, we have the definition that a word-root would be the part that doesn't change, the invariant part of a word. But what do "die" and "dying" have in common? What's the part that doesn't change? Just the "d." So is "d~" the stem, the lowest common denominator of these two words? But, wait... "dilemma" also begins with "d~," so would "dilemma" and "dying" be related somehow? Are all words that started with "d~" in the English language linked to "death"??!?? That's kind of absurd and not at all helpful, isn't it?
Root vs. Stem
To understand the difference between root and stem, I guess it's easier to see an example in practice. If you have a word like:
- Vaporization.
We can divide it into two morphemes:
- Vaporize + ~tion.
And we can divide vaporize in two morphemes:
- Vapor + ~ize.
We can't divide vapor further, so vapor would be the root. It's the root of vaporize, and of vaporization.
Meanwhile, since we can divide vaporize to get vapor, vaporize isn't the root of vaporization, vaporize is only the stem of vaporization.
Vapor is also the stem of vaporize, as the word is "the stem vapor plus the suffix ~ize."
Finally, there's the question of whether vapor is the stem of vaporization. I guess it depends on whether you care if it's a direct stem or not. It's not the stem if you want to look at it as three morphemes. "vapor + ~ize ~tion," but you could say it's the stem if you merge ~ize and ~tion into a ~ization suffix: "stem vapor + ~ization."
Similarly, in Japanese, sugo~ すご～ is only the stem of sugokatta すごかった, "was incredible," if you treat ~katta ～かった as a single, complex suffix. If you divide the suffix into two morphemes, ~ku atta ～くあった, then the stem of sugokatta is the ren'youkei sugoku すごく. But why would you divide it into two morphemes, when you could divide it into three: ~ku + ~aru + ~ta, and then the stem of sugokatta is sugoku aru すごくある.
- sugo~
- sugoku (sugo~ plus ~ku)
- sugoku aru (sugo~, plus ~ku, plus ~aru)
- sugokatta (sugo~, plus ~ku, plus ~aru, plus ~ta)
Although this level of detail is very interesting and helps explain some grammar, it's pedantic and useless most of the time. I mean, which one is easier to understand?
- The past form sugokatta is the stem sugo~ plus ~katta.
- The past form sugokatta is the ~ta jodoushi 助動詞 attached to the ~aru hojo-doushi 補助動詞 attached to the ren'youkei of the i-adjective sugoi, which is its stem sugo~ plus ~ku.
Stem vs. Base
The difference between stem and base is that the term stem is only used with inflection. All stems are bases, but not all bases are stems.
In inflected word-forms, a base is also called a stem, and occasionally this term is also used for bases of derived lexemes.(Haspelmath, 2002:19, emphasis mine)
Naturally, we'll need a a definition for inflection, then, if we want to make sense of this:(Haspelmath, 2002:268, 271, glossary)
inflection (or inflectional morphology): a part of morphology that is characterized by relatively abstract morphological meanings, semantic regularity, almost unlimited applicability, etc.
conjugation: (i) an inflection class of a verb; (ii) verb inflection in general.
declension: (i) an inflection class of a noun; (ii) noun inflection in general.
The "semantic regularity" noted in the definition refers to the contrast between inflection and derivation. While inflection always generates words with meanings you can guess from the suffixes (e.g. you may not be able to explain them, but you just know what ~ize and ~ation are supposed to mean), derivation does not.(Haspelmath, 2002:74)
For example, "notebook" is derived from the words "note" and "book." Since it's a kind of book, not a kind of note, I suppose the root of notebook is "book." This isn't an inflection, it's a derivation, so "book" is the base, but not stem, of notebook.
If it was an inflection, you'd expect to be able to add "note~" to all sorts of things, like notepaper, notecomputer, notesmartphone, notewall, notehand, etc., and everyone would know you're talking about a thing you use to take notes, which isn't really the case.
Similarly, sunflower, a kind of flower, not a kind of sun, isn't an inflection either: you don't say sunpillow and expect people to understand the pillow is sun-shaped.
You can't really predict what these derived words mean, because derived words, nouns, specially, are created to name things that already exist, so the meaning of the word depends on what the already-existing thing is supposed to be.
By contrast, inflections are used to change the meaning of the word to match a certain type of grammatical usage, so you can create a word that nobody has ever said before, but because the inflection's meaning is predictable, and stem's meaning is known, everybody just understands that the word means "it's the meaning of the stem, with the usage expressed by the inflection."
You could say something like "I lament the bamboozlification of our society," and even though I have no idea what I'm saying myself, one can reasonably understand that I'm talking about how our society has gradually become more bamboozled over time, whatever that means, because it's the meaning of the stem "bamboozle" with the usage of the inflection "~fication.".
A Warning About Roots
If you're just talking about grammar, and you aren't researching the etymology of words or doing something very linguistic about it, there really isn't much a point in using the term "root" at all.
In fact, if you use the term without thinking you may end up calling something a root that isn't even a root.
While it's very easy to tell long words full of suffixes have stems that are shorter words with fewer or no suffixes, there are cases a word that looks like it has only one morpheme actually has two, so they can't be a root.
For example: kokorozashi 志, "ambition," although spelled with a single kanji, is obviously composed of the word kokoro 心, "heart," and the ren'youkei of sasu 指す, "to point," affected by rendaku 連濁. So it's at least two morphemes, it can't be a root, even though it's spelled as a single kanji.
More complicatedly: kyou 今日, "today (this day)," kon'ya 今夜, "tonight (this night)," and kesa 今朝, "this morning," are all words which, very obviously, have some k~ morpheme that expresses the "this" meaning, so none of these words can be called roots, while, at the same time, it's impossible to extract the "this" morpheme from these words in modern Japanese. If you go back in time, when kyou was pronounced ke-fu けふ, it becomes clearer that the "this" morpheme could be ke~ け～.
Note: there was a time when the word was pronounced kyou, but was still spelled kefu けふ, despite that being an archaic pronunciation already. This stopped with the post-war orthographic reforms.
Even more complicatedly is the fact that a morphologically complex word isn't necessarily longer than its root. The definition of base we're using uses the term "morphological operation," not "agglutination." Morphological operations include, for example, changing a vowel of the base "win" to get "won," a scenario where both words are the same length, but one is the base of the other.(Haspelmath, 2002:23)
In Japanese, there's an argument that sugo~ isn't a root. The root is sugok~ すごｋ, as in sugoki すごき minus the ~i vowel, and you get sugoi through a morphological operation called consonant deletion, in which, literally, you delete the consonant ~k~ from the ~ki ending of sugoki, like you'd delete ~k~ from kaki-ta 書きた to get kaita 書いた (see i-onbin イ音便), and that's how you get the ~i ending. Similarly: yowaki 弱き would be the stem of yowai 弱い, kanashiki 悲しき would be the stem of kanashii 悲しい, and so on.(Fujiyoshi, 1982:88)
No comments: