MORPHOLOGY

3 definitions of word

orthographic, phonological, grammatical

argument against orthographic word

consistency issues (e.g. ice cream vs ice-cream) and requires a writing system which not all languages have

argument against phonological word

phonological processes which define it are language-specific, multiple processes may operate at the same time

argument for grammatical word

lexical integrity principle = rules of syntax can apply to entire words but not to internal parts of words

e.g. A: This is a cup of tea

B: What is this a cup of?

A: This is a tea-cup

B: What is this a … cup?

‘tea’ is part of the word ‘tea-cup’ in second sentence

argument against grammatical word

some cases of word components being separated in sentences (e.g. pre and post-revolutionary France), some processes target both phrases and parts of words

stem

base morpheme to which another morpheme is attached

root

simple indivisible base morpheme with descriptive meaning

affix

morpheme attached to a stem

how to identify a morpheme - fail

1. hypothesis - hear is morphologically derived from ear (hear = h + ear so h means use of)

2. test - is this supported by evidence from other words in the language? *h-eye and *h-nose

   1. conclusion - no other evidence for a morpheme h- so hear is monomorphemic/ a root

how to identify a morpheme - success

quickly, nicely, softly - -ly is a morpheme because:

1. sound shape - it appears in the same form -ly

2. meaning - changes the meaning of something having the property X to something being done in an X-ly way

3. distribution - it attaches to an adjective X to turn it into an adverb X-ly

   lovely does not belong because it does not conform to conventions 2 and 3

instances of the same morpheme

forms with the same sound shape, meaning and distribution in all their occurrences

complementary distribution

forms with the same function but different sound shapes may be instances of the same morpheme as long as their distribution is not the same so that one form occurs in an environment where the other never does (e.g. English plural morpheme -s, -en, nothing, -a in ‘cats’, ‘oxen’, ‘fish’ and ‘phenomena’)

evidence for non-segmental morphemes

ablaut (e.g. run vs ran), voicing alteration (e.g. cloth vs clothe), zero morph where it clearly contrasts with a non-zero variant (e.g. sheep vs sheep contrasting with dog vs dogs)

allomorphy

existence of different sound shapes for the same morphemes (e.g. imperfect subjective suffix can -ra or -se in ‘cantara’ and ‘cantase’ in Spanish), usually it is conditioned e.g. complementary distribution - 'an apple’ vs ‘a book’ ‘an’ and ‘a’ are allomorphs of the same morpheme, but conditioning can also be phonological, morphological, semantic and lexical

concatenative process

e.g. affixation - un+kind but this is conditioned so un- can combine with adjectives but not nouns

non-concatenative process

e.g. stress shift - recórd vs récord

mental lexicon

mental place that stores all words and their components or a mixture of both

morpheme lexicon

lexicon only contains simple, monomorphemic items (roots and affixes and idiosyncratic words) but not those whose meaning is predictable, all complex words are derived by rule

strict word-form lexicon

the lexicon contains all complex words regardless of whether they are predictable or idiosyncratic. The role of morphological rules is very limited the lexicon plays the same role as memory

moderate word-form lexicon

lexicon contains word-forms, morphemes and derived stems. Mixture of the previous two and allows for morphological rules

morpheme lexicon: advantages

1. parallel with syntax and phonology: speakers don’t memorise every sentence they utter (no need to memorise complex words either)

 2. maximally economical: no need to store every word form as a separate word, you can just add the correct ending onto a stem

morpheme lexicon: disadvantages

 - not all complex words can be understood by combining the meaning of their individual parts (non-compositional words)

 - not all complex words result from concatenative morphological processes

 - zero morphs and empty morphs are problematic for this kind of lexicon

strict word-form lexicon: advantages

 - word's meaning doesn’t have to be the result of combining the meaning and its morphemes

 - complex words resulting from non-concatenative processes are not an issue

 - helps explain uniquely morphological traits like lack of productivity

 - cognitively realistic (speakers are likely to remember a whole word if it is frequent, not just if it is unpredictable)

strict word-form lexicon: disadvantages

 - concatenative processes in richly inflected languages mean it is unlikely speakers will memorise all forms of all verbs

 - evidence that speakers see words of consisting of morphemes

 - languages don’t allow some sound sequences within the same morpheme, but okay across a morpheme boundary = speakers have to be able to acknowledge that there is morpheme boundary there

moderate word-form lexicon: advantages

 - both word-forms and morphemes can be lexical entries: represented as word-schemas, which interpret morphemes as morphological patterns

 - word-forms are still primary; morphological patterns have a secondary role

 - many complex words are listed in the lexicon, some composed from component parts when needed

moderate word-form lexicon: disadvantages

 - how is it decided which complex words are stored whole and which are broken down into their components?

 - how established does a word have to be to be memorised as a whole?

 - how are these stored words retrieved by speakers? By decomposition, or directly?

lexeme

word in an abstract sense representing the core meaning shared by word-forms

wordform

word in a concrete sense expresses the combination of a lexeme and a set of grammatical meanings

lexeme vs word forms

1. dictionaries have separate entries for lexemes (do, undo, outdo) but not for word forms of the same lexeme (do, does, did)

2. only word forms are ever used in language (we don’t utter lexemes only word forms of that lexeme)

endocentric compounds

compounds with an identifiable head (e.g. blackbird is a type of bird)

exocentric compounds

no identifiable head (e.g. pickpocket neither a type of pocket nor a picking activity rather person who…)

derivation

creation of a complex lexeme by applying a morphological process to a single, more basic lexeme, that does not involve several lexeme stems (e.g. logic, logical, illogical), often involves change of word class but it need not

non-concatenative processes in English

1. stress shift (recórd vs récord)

2. final consonant voicing (e.g. thief (N) -> thieve (V))

3. conversion/ zero derivation (dust (N) -> dust (V))

not always a straightforward distinction between derivation and compounding

e.g. peace-ful

compositionality

overall meaning of a complex lexeme is determined by the meaning of its component parts and the way in which those parts are combined (e.g. [[plastic truck] driver] and [plastic[truck driver]] = structural ambiguity)

establishing the order of processes

 - if we take off one portion, does the remaining bit exist? e.g. de + cipher-ment or de-cipher + ment, it is the latter because decipher exists but *cipherment does not

 - behaviour of affixes (e.g. -ment attaches to verbs to produce nouns and de-attached to verbs to produce verbs)

inflection

generation of the different grammatical forms of a single lexeme expressing morphosyntactic features (e.g. number, tense etc.)

3 types of cases

1. semantic case = bears a particular often spatial meaning (e.g. from inside the house in Finnish elative case)

2. structural case = based on syntactic structure (e.g. nominative, accusative)

3. lexical = depends on specific lexemes (towards takes dative nominals in Russian but up to takes genitive nominals in Russian)

exponence

the relation between a morpheme and its signified meaning

simple exponence

one to one relationship between form and function (e.g. sea-s = sea-PL)

cumulative exponence

a single formal element realizes a combination of morphosyntactic function (e.g. buon-o = good-M.SG, buon-i = good-M.PL, in Italian)

extended exponence

a single morphosyntactic is realized in different loci in the word (e.g. sol-d = sell.PST-PST)

form-function mismatches

zero morph, empty morph (a form with no function)

syncretism

the correspondence of difference sets of morphosyntactic features to a single word form (e.g. we are, you are, they are)

analytic end of the scale (language typology)

little morphology (all or almost all words are monomorphemic)

synthetic end of the scale (language typology)

morphology plays an important role; polysynthetic languages are extreme examples of synthetic languages

isolating languages

 - almost no morphology

 - a word correspond to one piece of meaning 

agglutinating languages

 - very complex words made up of multiple morphemes

 - correspondence between morphemes and functions/ exponents is one to one

fusional/ flexive languages

one exponent contains multiple morphosyntactic functions

polysynthetic languages

even whole propositions are expressed as single words

issues with typology

 - counting morphemes is not always straight forward (because of unmarked/ zero marked, what do to about cumulative exponence?)

 - languages can behave inconsistently across different word classes (e.g. Japanese verbs are synthetic but Japanese nouns are analytic)