Morphology
Morphology studies word structure, and and deals largely with morphemes.
It could be said to study the notion of possible words in a language.
This includes the regularities that e.g. allow us to understand words forms we have not seen before - which applies somewhat more to synthetic languages, in which the focus is on morpheme combination(verify).
Some terms
Morphemes are units of language that conveys meaning. A morpheme loses a particular meaning when divided further so are indivisible for a given meaning - though not necessarily indivisible at all.
Free morphemes are those that can stand alone (and be used as part of words).
Bound morphemes are those that occur only as part of words, such as english suffixes '-less' and '-ly.' (Note that in english, there is a free morpheme 'less' and a bound morheme in the suffix '-less')
A simple word is one consisting of a single morpheme.
A complex word is one consisting of a multiple morphemes, specifically a free one with various bound ones. This is often explained as a stem and affixes.
A compound (sometimes 'compound word') is one that combines multiple free morphemes into a single unit
A content morpheme is one that adds conceptual content/meaning to the sentence (or more content than function), often (the base of) nouns, verbs, adjectives, or adverbs.
A function morpheme is one that primarily adds relations between morphemes (or more function than content), including articles, prepositions, auxiliary verbs and such -- and affixes(verify)
lexical morphemes and grammatical morphemes usually mean much the same meaning as content morpheme and function morpheme (respectively). (verify)
Root, base, stem
The root is the most centrally meaningful part of a word,
- generally the last morpheme you have when you strip away when you remove affixes
Say, a base can be any morphological element to which others can be added.
- This is useful in that when we add things more than once, each addition has a clear base.
In the above:
- root leaves it open whether it's a complete word (e.g. free morpheme) or not
- That depends on the root, and the language.
- Bound roots are root those that are not complete words, that needs something more (usually inflection) to become usable.
For example,
- nomin- appears in nominate and nominee, frig- appears in frigid and refrigerator, just- in justify and justice, but not by itself.
- note that bases are more typically free
- I've avoided more language-specific definitions like those centering around derivation, inflection, affixes
- language teaching tends to use them because they're a lot more concrete, and good enough for a single language - and these will be the most common definitions you find out there
- but are more incorrect in more general theory, where it's a lot clearer to create more specific definitions, like 'inflectional base'
- I've avoided definitions like 'the unchangeable part of the word', because this can be vague about whether its morphographic or morphophonetic
Stem can be more confusing. As wikipedia points out[1], its meaning differs per language.
I've seen it used to mean base, (bound) root, complete word that can also act as a base, and a dozen other things.
Such meanings can be very useful e.g. within a course teaching a specific language, but in a generally analytic take, that linguistics usually takes, it is confusing at best.
You may want to avoid it, because if you don't, you run into conflicting definitions, talk about different topics entirely, or use circular definitions, habits differing between between different languages/language families, by people trying to force Latin theory directly on to language that work differently, between English teachers and linguists, between different linguists.
I've also seen definitions that seem to confuse together a bunch of different things, in the process of describing whatever the end result is for some language or specific inflectional effect.
In fact, you get some of that with just root and base.
Even with just base, questions still arise.
Can compounds be bases? If not, what's the term you'd use for that?
Does it apply to inflection and affixing only? Within what languages?
More discussion, Examples, possible statements
Affixes
Affixes are bound morphemes that are attached to lexemes to create form a word with a particular (marked) inflection, or to derive a (sometimes new) word from an existing one.
Affixes are bound morphemes (though some of them may also ), and are usually things that cannot be considered roots.
In various languages, suffixes and prefixes are the most common affixes.
There are others:
- A prefix attaches to the start of a word (e.g. unrealistic),
- a suffix attaches to the end (e.g. wonderfully),
- the less common circumfix to both ends (e.g. memperbaiki),
- and the infix somewhere in the middle (fairly rare except in relatively creative use like fanbloodytastic).
(There are more)
See also morphology, specifically the section on inflectional and derivational morphemes, and then specifically the affixes in the Enlish section.
Not all linguists are equally seeing lexically contentful beginnings and endings (e.g. -ology, neuro-, derm) and calling them affixes.
If you call them affixes, you can create words entirely of what you call affixes (say, neurology) and have to decide which one is the most rooty/heady.
Though if neither an affix, doesn't that make neurology a compond?
Also derm can be both beginning and ending, which is rare, so we don't like it.
So we play it safe, say 'bound morpheme', and/or quality what kinds of roots exist some more, and stay quiet when asked anything more.
Spoken or written?
Yes.
While examples of morphemes are often written, morphology does not explicitly have a take on whether morphemes are written or spoken, because the details to such may be deeply complex, and vary per language,
Morphology, morphographics, morphophonemics
Morphology in general is concerned with the combination and interaction of morphemes.
In many languages this interacts with writing and/or phonetics, so we have names for these areas.
In most applications you tend to focus on just one (text processing, speech processing), but in others it helps to be aware that looking at only one medium means ignoring some part of the morphology.
In many languages the morpheme matches well enough between sound and writing and meaning, so this is not much of an issue.
We tend to give examples as written text, which belies that morphographics (what happens in written form) is sort of our default.
Morphophonemics (or Morphophonetics) is the interaction between morphology and phonetics - which is often primarily the study of phonemic differences between allomorphs.
Allomorphs
An allomorph is the set of different pronunciations of a single morpheme.
Examples:
- variations in bound forms, e.g. leaf becomeing leav (as in leaves).
- the English pluralization affix can be realized with, most commonly, [s] or [z].
Compare with allophones.
Derivational and inflectional morphology
It seems traditional to make the distinction between derivational morphology and inflectional morphology, in many languages both referring to the use of function morphemes to create different word forms (usually using affixes).
However, the distinction between the two is somewhat idealized based on some properties it has, and there probably isn't a clear cognitive division. For some cases you can easily tell derivation and inflectional changes apart, for other cases even linguists will disagree on whether something is inflection or derivation/word formation. As such, it is a useful distinction to explain the differences involved, but it should not be thought of as a black and white distinction. The reality probably lies in more complex investigations of semantic and syntactic contributions.
(Note: in many languages there are many more content morphemes that can also be used as affixes, but this is (morpheme) compounding, not inflection or derivation. Of course, the the distinction between function and content affixes can in some cases be fuzzy too - the wider it can be used, the more it can be said to be a function morpheme)
You could say that derivational morphology is primarily lexical and does word formation, while inflectional morphology is more grammatical and only adapts words for use in different situations, answering questions like when and how many, but without changing the sense or function, and usually not the lexical category (POS) either. The meaning of the inflection is also usually very predictable, with minimal or no change to the sense (e.g. dog→dogs), only marking for a specific property (also a distinction you could argue over).
The distinction between inflection and derivation seems to exist mainly to point out that derivation does is a different thing, namely word formation: it creates words with different meanings/senses. Usually the sense of the derivation is similar and understandable, condider perhaps feel→feelings or sense→sensation. While similar and usually judged as sensible by humans, derivational changes are not always very predictable/analysable.
Derivational morphemes are not necessarily productive (can not be structurally combined with everything, or have erratic effect on meaning).
Inflectional morphology generally doesn't change the lexical category, derivational often does.
When both derivation and inflection are used in a word form, a language has rules of order and interaction, which are usually relatively strict.
In English; relatively common english affixes
Notes:
- You may generally add affixes to create longer words and remove affixes to get more basic forms. Related notes:
- Particularly additions will often be understood, even most that are not common or dictionary-accepted
- removing affixes won't necessarily yield existing words, and not only because you get a bound root or one that you never see alone (some words are introduced this way, though; see back-formation)
- affix changes won't necessarily yield things from the same root, or may form a homograph that also matches some other sense
- specific affixes tend to work only on specific types of words. For example, a squeegee is one one who got sqeeged and Jeffery is not a place that Jeffs and a theory is unrelated to the; -ee works only on verbs and -ery only on (non-name) nouns and on adjectives.
- just lexical category filtering is not enough to make strong assertations with senses. For example, the noun miner plus the suffix al yields mineral, which does not mean 'related to miners.' Similarly, medicate does not have the meaning 'to medic someone'.
- it's not unusual for a word to have two or three affix changes, and inflectional and derivational can happen side by side.
- A few cases (e.g. -ing), the same morpheme act in inflectional and derivational morphology. Consider work→working (inflection) and build→building (as in a building), which is derivation.
- Occasionally you get homographs with different senses.
- The distinction between affixes and morphemes may be vague. For example, dia- occurs almost exclusively in Greek loanwords and has a fairly well-settled meaning perhaps in part becaise of its limited use. Of course, the whole idea that an affix is functional breaks with a god amount of (relatively less common) specific affixes that are could also be seen as (some degree of) content morphemes.
- English used to be a highly inflected language, but modern English morphology is mostly derivational. In fact, the list of inflective suffixes below can be considered a complete list for current American English.
- There may be choices affected by other words. For example, there are various prefixes that have a not/opposite meaning. Usually only one or a few are possibly on a specific word
Some more cases to consider:
- manual technically comes from manus+al, but many might consider manual a root
- The word ran has nothing to do with the word random, and so on.
- The pre- prefix commonly appears with bound roots
Inflective suffixes
- -'s suffix: posessive noun (genitive)
- -s/-es suffix: plural noun
- on a singular noun -> plural noun (e.g. cats)
- on a verb -> (3rd singular) present (e.g. She trips)
- -ed/-en suffix: past verb tense (e.g. eaten, has closed)
- -ed suffix: participle marker (e.g. he danced)
- -ing suffix: (present) participle marker)
- -er suffix: comparative adjective
- -est suffix: superlative adjective
Derivational prefix examples
- be- prefix
- verb->verb
- noun->(verify)
- de- prefix (away)
- (verify)
- dis- prefix (sense of not/opposite) (note that the s falls away if prepending before a letter b, d, l, m, n, r, s, v, and occasionally in front of g and j)
- en- prefix
- (verify)
- in- prefix (sense of not/opposite)
- adjective->adjective, (verify)
- im- prefix (sense of not/opposite)
- adjective->adjective, (verify)
- mis- prefix (sense of not/opposite)
- verb->verb
- pre- prefix ()
- verb->verb
- re- prefix (sense of again)
- verb->verb
- un- prefix (sense of not/opposite)
- verb->verb
- adjective->adjective
- (verify)
Derivational suffix examples
- -al suffix (relating to, pertaining to)
- noun-> adjective (verify)
- -an: see -ian
- -ant suffix (performing agent)
- verb->noun
- -able/-ible suffix (relating to) (See also ible, able)
- verb -> adjective
- -ance/-ence, -ancy/-ency suffix (act, state, quality)
- verb->noun
- -ary/-ery/-ory suffix (relating, quality,place)
- noun -> noun
- adjective -> noun
- adjective -> adjective (verify)
- -ate suffix (cause, make)
- noun->verb
- adjective->verb
- -dom suffix (state, quality)
- noun -> noun
- -ee suffix (receiving party. Applies to transitive verbs that select for animate objects)
- verb->noun(verify)
- -ess suffix (female)
- noun->noun
- -ese suffix (inhabitant, native of)
- noun->noun
- adjective->adjective
- -ey suffix
- ->adjective
- -en suffix (become, make)
- adjective -> verb
- (verify)
- -ence suffix: see -ance
- -er, -or suffix (actor, one who ...)
- verb -> noun(verify)
- -ful suffix (having, filling)
- noun->adjective
- verb->adjective(verify)
- -fy suffix: see ify
- -ible suffix: see -able
- -ian/-an suffix (inhabitant of, pertaining to)
- noun->adjective
- -ic suffix (characteristic of)
- noun -> adjective
- (verify)
- -ical (see also -ic, -al, and ic, ical)
- ->adjective
- -ity suffix (condition, quality)
- adjective -> noun
- -ing suffix
- verb->verb
- noun->verb (verify)
- -ive suffix (tending to, doing)
- verb -> adjective
- -ily/-ly suffix
- adjective -> adverb
- -> adjective (sometimes)
- adjective -> adverb (verify)
- -ise/-ize suffix
- adjective -> verb
- noun -> verb
- -isation/-ization suffix (change into, process), a combination of -ise/-ize and -ate, -ion suffixes
- -ish suffix
- adjective->adjective
- noun->adjective
- (verify)
- -ification suffix (change into), a combination of suffixes (-ify, -ic, -ate, -ion)
- -ion suffix: see -sion/-tion
- -ish suffix (nature, resembling)
- (verify)
- -ism suffix (system, characteristic)
- noun -> noun
- adjective -> noun
- -less suffix (absence)
- noun -> adjective
- -like suffix
- (verify)
- -ly suffix: see -ily
- -ment suffix (state)
- verb -> noun
- adjective -> noun(verify)
- -ness
- Adjective -> Noun
- -ous suffix (having, full of)
- noun->adjective
- -or suffix: see -er
- -ship suffix
- noun -> noun (verify)
- -tion/-sion suffix (act of)
- verb -> noun
- -y suffix (having, full of)
- noun->adjective
- yze/yse: see -size/ize
- yzation/ysation: see -ization/isation
Morphographic details
Even when there is no sound change (or sometimes to prevent it), you may have to do something more than just adding an affix. This often consists of adding or removing a letter or two, and not unusually doubling a consonant.
Such phonological details are not all to predictable in English in general, and it only gets a little better when looking only at suffixing (or affixing in general, but prefixes tend to be simple separated syllables).
- vowel end often means removal of the vowel (verify)
- vcvc ending, cvc ending (often indicating doubling(verify))
- cc endings (usually doesn't involve doubling)
- stress (stress on/near the end indicates doubling(verify))
- ending with the letter c often means a k is added.
See also
- http://en.wikipedia.org/wiki/Derivational_Morphology
- http://www.indiana.edu/~hlw/Derivation/derivation.html
http://www.ling.upenn.edu/~rnoyer/dm/
The isolating / synthetic / polysynthetic scale
(Note that terms are more tendencies of morphological behaviour, less than than strict language typology. They are pragmatic, not precise. For example, 'analytic' is often contrasted with 'synthetic', yet those two concept are not considered gradations on the same scale.)
Isolating languages
- those where most morphemes are free morphemes and usable independently
- In most, there are few or no affixes or other bound morphemes. Generally not a lot of compounds, with footnotes.
- Examples include Mandarin
Analytic languages, depending on who you ask,
- may be a near-synonym for isolating languages, or
- may have a definition that mentions how it expects information to be coded (e.g. using particles, posessive markers, prepositions, articles).
- Just how similar that makes it to isolating, and how implicit/explicit that is, varies accordingly.
Synthetic languages
- are those that rely on morpheme binding
- in the form of inflection and declension, or fusion, or agglutination.
- mainly contrasting with isolating(/analytic), and usually we use a more specific term (see below)
- (These languages also often have agreement patterns that help settle references and sentence structure.)
Polysynthetic languages, are synthetic but more so.
- Due to large amounts of bound meaning, words/compounds could contain as much information as what other languages do in clauses or entire sentences
- (though this stretches the concept of a word (and sentence), which are arguably not very applicable to those languages in the first place.)
- Examples include Mohawk
A distinction mainly made in the context of synthetic and polysynthetic context (but parts of which also applies to 'merely' synthetic languages) comes down to whether affix morphemes inflect for one or for more than one grammatical meaning at a time:
Fusional languages (sometimes amalgamation) are those where an added morpheme can mark many things at once.
- Say, in Latin amo, -o indicates mood, active voice, first person, singular, present tense
- Examples include Latin, Spanish (verify)
Agglutinative languages basically means that each affix adds only one such grammatical meaning at a time
- For example, in Turkish anlamadım we have a root (anla-, to understand), marks it as a negative (–ma), definite past (–di), and first person (-ım), which combines to mean 'I did not understand'
- Examples include Finnish, Estonian, Hungarian, Turkish (verify), Tamil
Notes:
- Polysynthetic languages are more often agglutinative than fusional
- (Which is probably nicer for learners, as each marking affix can be understood individually. Fusional is more succinct but may have hundreds of distinct affixes to learn)
- These descriptors are rarely pure distinctions
- as as evidenced by "so what are languages like German? Persian?" being hard questions
- e.g. people can't quite agree on where German falls in the above
- none of this really addresses things like purely compositional compound nouns such as the ones German and some other languages are fond of (but are allowed happen in others), e.g. Studentenkrankenversicherung ('student sickness insurance')
- you could consider it a special case of agglutination but it's stretching the concept
These aren't simple, singular language descriptions.
For example, consider that Japanese is agglutinative and synthetic in its verbs, but analytic in its nouns.
Most Indo-European families are moderately synthetic, tending to use one or two affix morphemes to mark at least one or two of number, person, gender, and case.
This is a matter of degree.
Compared to Spanish, for example, English is relatively analytic.
Yet antidisestablishmentarianism is six morphemes of English agglutination - that's not a compound.
But also, that comparison depends on whether you consider Spanish a synthetic or polysynthetic language, which means it's a slippery statement when you contrast with something that lies on a scale.
Point is, few Indo-European languages do complex inflection.
A few make long compounds, sure - e.g. German and to a lesser degree Dutch are fond of this (Kraftfahrzeughaftpflichtversicherung, Vervoerdersaansprakelijkheidsverzekering) - but compounds aren't inflection.
Chinese, Thai and Vietnamese are examples of fairly analytic languages, but there aren't a lot of them, and note that even many logographically geared systems like Chinese (one of the best examples of isolating languages) are often not strictly isolating, just mostly - it does have bound morphemes.
There are also relatively few languages strongly on the other polysynthetic end, and also you run into the problem that 'morpheme' and 'word' aren't clear distinctions as they are in the languages that linguistics started with an is biased towards having descriptive words for.
Measuring synthesis
One metric of where to place a language is to figure out the morpheme-to-word ratio - how many morphemes there are in a word, or rather, all words.
If that ratio is 1:1 for most (formed) words in a language, the average of this ratio for the words in the language will be near 1.0. If most words in a language have more than one morpheme, it will be higher.
This will often be a good indication of how much synthesis a language tends to use, and a basis of calling things isolating languages, synthetic languages, polysynthetic languages, or to perhaps have a little more nuance in such a scale.
As an aside, languages that are more synthetic than analytic tend to have more regular patterns to their inflections, and fewer exceptional constructions. Less regular languages tend to be the analytic ones.
You may see reference to inflected languages' or inflectional language (seemingly a near-synonym for fusional), but this is often considered a confusing term from an obsolete typology.
See also:
- http://en.wikipedia.org/wiki/Morphological_typology
- http://en.wikipedia.org/wiki/Isolating_language
- http://en.wikipedia.org/wiki/Synthetic_language
- http://en.wikipedia.org/wiki/Polysynthetic_language
Word formation
Word formation is typically a primarily morphological exercise
Computational morphology
Computational morphology often focuses on writing, and so can usually be described as morphographemic systems, and many may focus on generation or analysis but not necessarily both.
Observations you can use in such systems:
- Affixes generally apply to words in a few lexical categories
- additional filters may apply (e.g. -er and -ee affixes select for transitive/distransitive verbs that work on animate objects/subjects
- verbs may be
- Most affixes create a word in either in a specific new lexical category, or the same as the one it applied to
Unsorted
Agglutination
Agglutination refers to combining morphemes into a single syntactic feature. ...and is no more specific that that.
Agglutination is a useful term e.g. when describing and contrasting different forms of word formation between languages.
In a lot of languages agglutination works out mostly as affixes and then usually suffixes, but the term 'agglutination' historically has had more varied meaning[2].
There seems to be no single clear definition, so the term seems to mostly come up in contrasts, e.g. when quantitative linguistics points out contrasting tendencies such as:
- isolating languages (little to no affixing) with
- agglutinative languages (affix tends to add one piece of meaning) and
- fusional languages (affix can mark many things at once).
Note that none of that describes e.g. German's tendency to make compositional compounds more and longer than most other languages.