Návrat na detail prednášky / Stiahnuť prednášku / Univerzita Pavla Jozefa Šafárika / Filozofická fakulta / Britské a americké štúdiá

Morphology Lectures (morphology_lectures_-_bas_08-09.doc)

MORPHOLOGY & ITS BASIC UNITS

1. J. W. Goethe – die morphologie – formenlehre
2. firstly introduced in Biology; 19th century – in linguistics
3. studies shapes and internal structures of words – correlation of form and meaning of the word (W)
4. grammar = morphology + syntax + phonetics & phonology
5. M was considered to be well researched -> lesser attention (than syntax)
6. 70s – M development, increased interest, innovations
7. fundamental concepts of M – word & morpheme (least consistently used in M)

Status of Word

1. defining W syntactically: smallest unit of syntax, criterion of the minimal free form (Bloomfield)

- a potential diagnostics for wordhood (can it stand alone?) – not universally reliable (whose book is it? – my* (should be "mine"))

2. semantically: semantic entity, a linguistic unit (LU) of a single/more meaning/s

- frequently word boundaries do not coincide with meaning boundaries (criminal act – criminal lawyer)

3. phonologically: domain of stress assignment – not an absolute generalization (The hot dogs ran for the lake – 7 words, 4 stresses)

Content Words & Function Words

1. CW = lexical word – refers to objects, events, abstract concepts... – open class
2. FW – best characterized by its grammatical function – modals, articles, conjunctions... – closed class

orthographic W – unit, which, in print, is bounded by spaces on both sides (words separated

        by gaps)

lexeme – abstract unit unifying several word-forms (WF)

8. lexemes – in print in capitals (AND)
9. lexeme vs. lexical unit – Cruse (1986): LU – union of a lexical form in a single sense, L – family of all LUs

paradigm – set of all inflected forms that a lexeme assumes

grammatical W – defined by its place in a particular paradigm – different gram. function (he

        walked home – he has walked home)

idioms – also listed in dictionaries (but are not words) – share with lexemes the quality of

        having dictionary entries

listeme – listed in dictionary (lexemes, idioms, phrasal verbs...)

Bauer (2003) – word-forms (including orthographic Ws)

                         – lexemes

                         – grammatical  Ws

Matthews (1974) – phonological (or orthographic) Ws

                              – grammatical Ws

                              – lexeme

DiSculio & Williams – 4 notions of W

1. morphological object – constructed out of morphological atoms, i.e. morphemes, by processes of affixation & compounding
2. syntactic atom – indivisible building blocks of syntax and thus treated as indivisible into morphemes
3. listeme – W as a listed object, linguistic expression memorized and stored by speakers
4. phonological W

Elements Smaller than Word

morpheme – smallest, indivisible unit of semantic content or grammatical function from which Ws are made up; smallest linguistic sign (LS) – form + meaning

10. Jan Baudouin de Courtenay (19th C)
11. phonological difference correlates with semantic difference (boy/boys), (the boy goes/the girl goes)
12. the difference in lexical meaning (she plays/she played)
13. the difference in grammatical function

morph – physical morph representing some morpheme, realisation of morpheme

14. in writing – morphemes enclosed in { }, morphs separated by decimal points (great.ly)

   LEXEMES                 made up of 1/more                         MORPHEMES

   realised by                                                             realised by

WORD FORMS                made up of 1/more                             MORPHS

15. if different morphs represent the same morpheme – allomorphs of that morpheme (/id/, /d/, /t/ - past tense)
16. a set of morphs is classified as allomorphs of the same morpheme if they are in complementary distribution
17. CD – if they represent the same meaning or serve the same grammatical function and are never found in the same context

DERIVATIONAL & INFLECTIONAL MORPHOLOGY

1. differences between D & I M – any clear-cut boundary between them?
2. prototypical cases of derivation & prototypical cases of inflection
3. different degrees of D & I
4. hypothesis – no clear-cut boundary between D & I – cross-linguistic comparison (one phenomenon of D in one language can be of I in another language)
5. criteria to distinguish D & I (always some exceptions):

1. DM is based on semiotics

6. DM produces new linguistic signs – we name new objects (any substances, actions, qualities, circumstances etc.)
7. these objects (extralinguistic reality) result from our cognitive (intellectual) activities
8. cognitive activity –> extralinguistic reality –> LS
9. IM – no new words, we remain within one LS (practically this does not always work – is it a new LS or not?
10. e.g. deminutive otec –> otecko – ist it a new LS? – denotes only a new aspect of the object
11. pluralactional verbs – express plurality in some langs (V + inflectional morpheme – distributive action/iterative)
12. semiotic function of DM is mainly theoretical
13. semiotic function (DM) vs relational function (IM – expresses relations between words and sentences – he readS expresses relation of he + read)

1. IM is affected by syntax, DM affects syntax

14. the syntactic structure affects the form of the word (they ARE crying – determined by they)
15. also some exceptions
16. general (Adj) – has syntactic function – attribute/predicate
17. generalize (V) – different function – not attributes, compliments... = new word

1. IM is obligatory, DM is not (regarding syntax)

18. e. g. hovorím s chlapom, not s chlapov (I) -- > obligatory
19. thief (moneme), stealer (complex, derived word) – you can use any of them (no rule)
20. the poor, the rich – no plural; dead, round (no deader, rounder) --> exceptions

1. competition

21. no competition between "chlapa" (G) and e.g. "ženy" (G) --> "chlapa" in genitive
22. "-ness", "-ity" – suffixes with the same meaning (to create abstract nouns)
23. -er, -ist, -ian, -man,  - same meaning (to create agent) --> competition between them – we can use any of them (to a certain degree)
24. the selection depends on the age, education, profession, linguistic background, gender of the speaker (when coining new words)
25. more options in DM (chessist, chessman, chess player)

1. uniformity

26. greater in IM – consequence of competition & obligatory criteria
27. sth is obligatory, no other choice --> you use the prescribed form --> uniformity
28. if you have a choice (of rules, affixes) --> there are agents ending in er/ian/ist/etc. – not all agents end in –er --> er-system has gaps
29. rules that are almost completely uniform – ing forms

1. R. Beard – 1995 – at least in Indo-European languages – common morphological categories (44) – in individual lang. can acquire D or I manifestation – distinction between abstract and concrete categories is not clear

30. semantic relevance & amount of semantic change – criteria proposed by Bybee
31. SR – a meaning element is relevant to another meaning element if the semantic content of the 1st directly affects/modifies the semantic content of the 2nd
32. tree – trees – s does not affect the semantic context of tree – difference only in quantity
33. dog – doggie – semantic content is slightly changed (small dog)
34. write – writer – semantic content changed – derivation with substantial change of SC

35. DM – change of the semantic content of the base
36. the greater the difference between the meaning of the derived word & the meaning of the base, the greater the likelihood that the affix is derivational

1. IM is morphosyntactically more transparent than DM

37. morphosyntactically = morphemes + their meaning
38. tree-s – morphosyntactically transparent = plural of tree – principle of compositionality
39. analytical langs. – minimum of inflection
40. G – chlapa, A – chlapa --> -a for 2 cases --> we do not know what –a means (PoC not applicable here)
41. syncretism – 1 form for more grammatical meanings
42. -a = 3 gram. meanings – M gender, sg, G --> cumulative exponent (1 morph. – 3 meanings)

43. airs – unnatural manners in order to impress – air has no plural, air+s = unpredictable
44. damages – money to be paid for causing damages = unpredictable
45. works – moving parts of a machine = unpredictable

1. protoypical IM does not change the word class

46. write (V) + er = writer (N) --> DM
47. word class – one aspect of categorial change
48. S. Scalise – also changes within the word class – we should go into the word class, not above it

                                man                                        manhood  –  both nouns

                         - abstract                                        + abstract

                        + countable                                           - countable

1. morphological rules of DM can be recursive (in contrast to IM rules)

49. tree –s –s  --> treeses --> no sense (IM)
50. general –ize –tion –al (DM)
51. institute –ion –al –ize –ion –al --> makes sense (DM)
52. we can repeatedly use morphemes in DM
53. IM – one rule applied once only
54. Jaqam (lang.) – N>V>N>V>N...
55. West Greenlandic (Eskimo) – 400 productive affixes – can be combined with each other recursively --> long words (correspond to our sentences)

1. IM is typically organised in paradigms

56. paradigm = set of all inflected forms of a word
57. paradigmatic organisation in DM is much weaker
58. if you produce a new word, it is classified as a paradigm (has rules)
59. DM – list – lístok – lístkový – lístkovitý
60. repeated patterns of regularity – květ – kvítek – kvítkový – kvítkovitý

1. derived words are likely to be stored in memory (inflected word forms unlikely)

61. lots of psycholinguistic research here
62. in memory we keep patterns, not words (sing + s = 3rd sg --> automatically we can produce other 3rd singulars)
63. also patterns for regularly derived words – [V + er --> NAg]
64. we remember idiosyncratic cases (unpredictable) – eg irregular plurals, words like transmission (unpredictable meaning "gear")
65. we have to remember cases that can't be formed according to a rule
66. idiosyncratic cases prevail in derivation

1. I affixes have more peripheral position than D affixes

67. general – ize – s                                        driver – er – s

                      D           I                                                   D    I

68. exceptions:       betterment (better - form of good)                                                

                             I     D

69. unhappier  - un [happier] – morphological analysis (correct)        = bracketing

                         [unhappy] er – semantic analysis                                     paradox

MORPHOLOGY / PHONOLOGY   INTERFACE

1. interface – way in which M and P affect each other
2. includes cases when P uses M information or vice versa
3. these interactions and grammar describing them = morphophonology or morphophonemics
4. allomorphy – the English past tense suffix –ed --> [d] (blamed, triggered), [t] (jumped, itched), [id] (aided, loaded)
5. the underlying (basic) past tense morpheme is [d] – in 2 unlike environments (after all vowels and specific consonants)
6. [t] – result of assimilation (preceding consonant is voiceless)
7. the vowel in [id] is epenthesis
8. assimilation – occurs when one segment takes 1/more phonetic characteristics (nasality, voicing) of the another one
9. progressive assimilation – characteristics spreads forward
10. regressive assimilation – characteristics spreads backward
11. epenthesis – process that inserts a segment in a given environment

                        drama – dramat-ic  –  dramat-ist

                        Plato  –  platon-ist  –  platon-ism

                        (short)     (longer)        (longer)

12. long form is the underlying form – to be used fot non-native (Latinate) suffixes
13. short form – used before native (Germanic) suffixes and with prefixes
14. plural of drama = dramas, not *dramats

Prosodic morphology

15. deals with interaction of M and prosodic structure
16. prosodic structure – concerned with the timing units of lang. (word, syllable, vowel length)

1. root-and-pattern morphology

17. esp. in Semitic languages, root consists of a series of consonants and derived/inflectional forms are created by superimposing the root on a vocalic pattern

                        malax (he reigned) – root m l x – template _a_a_

2. reduplication

18. morphological process that repeats all/part of a given base
19. determined in phonological terms, used for D and I

                        kaldi (goat) – kal-kaldi (goats)

                        pűsa (cat) – pűs- pűsa (cats)

Linguistic exaptation

20. diachronic; occurs when phonological material takes on new function unrelated to its original and obsolete function
21. Lass (1990) – "linguistic junk"
22. Rudes (1980) – "linguistic left-overs"
23. the development of the verbal suffix –esc in Romance:
24. in Latin the suffix –sc attached to sequences of verb stems plus theme vowel (determining the inflectional class) to form the inchoative aspect (an action about to begin)

                                paleo (I am pale) – palesco (I begin to pale)

                                amo (I love) – amesco (I begin to love)

25. inchoative suffix stopped to be productive (in time)
26. theme vowel came to be segmented as one invariable form
27. -esc in Romanian & Rheto-Romance + -ice in Italian

28. Lass' model for the future development of ling. junk -esc/-isc:

1. could disappear completely
2. kept as "marginal garbage"
3. take on new meaning and function (this happened – it has distinctive function)

Levelling

29. diachronic process of simplification & regularisation of M in which an irregular or minority pattern changes to a more common pattern
30. sound alternations that do not signal important differences in meaning are often eliminated

Prehistoric Latin                         Old Latin                                Classical Latin

N         colo:s                                       colo:s                                   color

G        colo:s - e                                 colo:r - is                                   colo:r - is

D        colo:s - ei                                colo:r - ei/e                                colo:r - i

A        colo:s - em                        colo:r - em                                 colo:r - em

Ablative        colo:s - i                  colo:r - e                                colo:r - e

31. due to rhotacism intervocalic /s/ became /r/

32. analogy – the force which allows the speaker to create a new construction based on already-

known parallel form

33. children often say *comed, *maked (by analogy with regular verbs, e.g combed)

MORPHOLOGY / SYNTAX  INTERFACE

1. I expresses morphosyntactic information (syntactic information expressed morphologically)
2. this includes the abstract syntactic categories of tense, aspect, number and case
3. specific values for these categories, eg past, plural, are generally referred to as morphosyntactic features (Matthews) – the feature is morphological but required by syntax

Exponence

4. coined by Matthews – refers to realisation of morphosyntactic features in inflection
5. in an one-to-one relationship between form and meaning we speak of single exponence
6. if more than 1 morphosyntactic feature maps to a single form, we speak of a cumulative exponence (eg Russian stol-a expresses G & singular)
7. if a single morphological feature is realized simultaneously on more than 1 form, we speak of extended exponence (sell - sold = e-->o, +d)

Context-free inflection

8. a simple directional mapping between a morphosyntactic feature and a particular phonological string (eg -ing)

Context-sensitive inflection

9. the realisation of morphosyntactic features varies (eg ablaut – ran; suppletion – was, went; zero – hit, cut, sent, lent...)

Inherent inflection

10. basic to the word and it does not have to be assigned to it (eg gender of the nouns)

Assigned inflection

11. a result of government or concord with another element in an utterance (eg case)

Concord = agreement

12. occurs when 1 element in a sentence takes on the morphosyntactic features of another element (he fishes = subject-verb concord)

Government = rection

13. refers to a situation when 1 word in a sentence determines the form of another (eg case assignment by verbs – vidím chlapa or he gave me a ring – not he gave I a ring)

Inflectional categories (by Anderson)

1. configurational properties

14. the choice of a particular I is determined by the place occupied by a word in a syntactic config. structure (eg direct object of a verb must be in the A cases)

2. agreement properties

15. determined by the characteristics of another word/s in the same construction (eg if an adjective modifies a singular noun, it must be assigned a singular affix whose form depends on the form of the affix in the noun it is modifying – pekné dievča, not pekný dievča)

3. inherent properties

16. must be accessed by agreement rule (eg the gender of a noun – the gender of French or German nouns determines the gender of the adjective that modifies it)

4. phrasal properties

17. belong to entire syntactic phrase but are morphosyntactically realised in one of the words of that phrase (eg genitive marking 's in Eng. – the Mayor of the Lancaster's car)

Verbal inflectional categories

1. Inherent verbal properties

18. tense, aspect, mood, conjugation classes – add extra specification to the event, state, process or action indicated by the verb – it will be referred to as the predication

1. tense – indicates the time of the predication in relation to a particular moment (whether the event happened before the moment (past), at the same time (present) or after it (future))
2. aspect – indicates whether the action/process/etc. is complete or in progress (imperfective = progressive or perfective = completed action)
3. mood – describes event in terms of whether it is necessary/possible/permissive/desirable (you must – necessity, you may – permission, you can – possibility)
4. conjugation – denotes inflectional classes for verbs; in many langs V belong to a number of distinct morph. classes called conjugations and verbal paradigms are normally described in terms of conjugation

1. Agreement verbal properties

19. agreement markers determined by the characteristics of some other word in the same construction
20. in Eng. – the V agrees with the subject in number (he walks – they walk)
21. the only V with extensive agreement is be (am/are/is)

Inflectional categories of nouns

1. Inherent categories of N

22. gender and number

5. number - in many langs – 1 and more than 1 (woman – women); other (Greek) – 1, 2 (dual) and more than 2

23. other – further distinction (paucal = a few)
24. G – obligatory
25. Chinese – no inflection for number

1. gender – Eur. langs – nouns referring to animate individuals are usually masculine or feminine, inanimate are usually neuter

26. gram. gender does not always match biological sex (das Mädchen)

1. Agreement categories of N

27. within a noun phrase agreement rules copy an inherent feature of the noun (G+N) on to other words such as articles, numerals and adjectives in the construction
28. French – La petite fille chante – The little girl sings

                     Le petit garcon chant – The little boy sings

1. Configurational categories of N

1. case – many langs in which subjects are in Nominative case and objects in Accusative = Nominative-Accusative lang. (Eng., Slovak; I saw him)

SEMANTIC APPROACH TO MORPHOLOGICAL ISSUES

1. this approach is not typical for US linguistics (the approach is purely formal, disregarded semantics – meaning is difficult to capture)
2. later – generativists – generative semantics
3. US ling. – Robert Beard (better known in Eur.), Joanne Bybee

Joanne Bybee

4. morphologist & typologist; analysed 50 langs – tried to explore relation between DM & IM
5. 3 important notions as criteria – semantic relevance, the amount of semantic change and generality of the meaning of the morphological items
6. kill – contents semantic component CAUSE
7. walked – PAST
8. come to know – INCHOATIVE (beginning of the action)

1. cause – identified semantically (in the meaning)
2. past – expressed within a morpheme –ed
3. inchoative – in a phrase (syntactically)

--> different ways of expressing semantics

9. emphasises obligatoriness of I markers (D are not obligatory)

Semantic relevance

10. a meaning element is relevant to another meaning element if the semantic content of the first directly affects/modifies the semantic content of the second
11. čítal – čítala --> -a does not affect the stem, the action is the same – semantic relevance of –a is almost zero
12. čítať – čitateľ --> -eľ affects the meaning (action --> person) – substantial semantic change
13. if an element is of high semantic relevance --> it is derivational
14. if an element is of low semantic relevance --> inflectional

Class changing derivation – word changes its word class

Class maintaining deriv. – same word class (boy --> boyhood – concrete N --> abstract N)

                        he reads a newspaper

                        he is reading a newspaper – class changing D (derived from read); -ing –

     semantic relevance almost zero (action --> process)

        there is a continuum  -->                    0 ___________________ max

                                                  inflection                        derivation        

Meaning generality

15. consistent application of meaning across a particular category of roots with the same semantic effect
16. chlap – all the words declined according to this have the same case endings (only a few exceptions)
17. verbs – regular (1sg = -m, 2sg = -š) – the applicability of the morphemes across the root (eg no matter what the root of the noun is, it is declined according to chlap)
18. all the roots are controlled by the same principle
19. D morphemes are barely general, there are many gaps – higher competition

Robert Beard

20. neglected by US linguists; several books
21. new theory – lexico-semantic morphological theory – lexeme-morpheme base morphology (LMBM)
22. basic principles – study of Indo-European categories (synchronic & diachronic approach)
23. D categ. rules and principles are common to all Indo-Euroropean langs
24. they are of universal validity and applicability
25. they are expressed differently in various langs
26. those universal categories and functions must be separated from morphemes which mark them
27. we have semantic categories operating on different levels than their morph. realisation
28. he identified 44 semantic categories – may be expressed derivationally or inflectionally in different langs (no border between D & I)
29. in the beginning of Indo-European langs there was no D or I, only 1 category
30. any attempt to separate D and I is artificial

31. Jan Baudouin de Courtenay (PL) – puts roots, affixes (D&I) into the same component – morpheme = single morpheme hypothesis

32. Bloomfield – transferred this idea of linguistic sign from word to morpheme – sign base morpheme hypothesis

33. morphemes/affixes are in lexicon with the words (general approach) = lexical morpheme hypothesis

Robert Beard

1. critical with regard to former approaches
2. lexeme-morpheme base hypothesis – distinguishes between lexical morphemes (lexemes) and grammatical morphemes (morphemes)
3. lexemes – free units, morphemes – bound; clear boundary between them
4. lexemes – open classes (new words can be added)
5. this correspond to Aristotle's approach (major and minor word classes)
6. bound morphemes (affixes) – same status as conjunctions, interjections etc. (grammatical function) – D & I morphemes (no meaning, just function)
7. lexemes – direct connection to extralinguistic reality, refer to objects, are signs, have meaning
8. lexemes are phonologically represented (morphemes need not to – zero/empty morphemes)
9. lexemes have categorial meaning, morphemes represent meaning, but do not carry it, they just spell out (express) the categorial meaning of lexemes
10. morphemes refer to universally available categories --> certain number of categories common to Indo-European langs – represented by I & D morphemes or syntactically
11. I category can be realised differently in different langs (eg comparison expressed with affix or syntactic devices) – this pertains to all 44 categories proposed by Beard
12. all Indo-European langs can make use of these categories and have their own ways of expressing them
13. eg category Anterior (sth is before sth else in terms of time) can be expressed by preposition before or prefix fore- or prefix pre-
14. category Intermediate – syntactically --> preposition between, derivationally --> prefix inter-

B. Šimanek

15. the basic set of lexical D categories is routed in the fundamental concepts of cognition
16. categories expressed differently

17. D morphemes are separated from I morphemes – separationist hypothesis
18. I spells out categories in syntax, D categories in lexicon

category Agent (Subject – the bearer of action)

        The baker bakes bread.

bake = lexeme

category of object – bake + er

19. certain relation between bake & baker
20. one category (eg Agent) can be expressed differently – derivationally –er, inflectionally – position in the sentence (at the beginning of a sentence – subject)
21. Beard distinguishes between derivation (cognitive process) and affixation (spelling-out process – phonological representation without meaning)

22. lexemic extensions are regular, productive, abstract rules producing regular semantic elements without any reference to affixation
23. Beard rejects Chomsky's opinion that derivation is irregular --> D rules are highly predictable
24. Ag --> -er, -ist, -ee, -man, -ian,... --> a number of affixes for one category
25. also one affix can spell out different cognitive categories (-er – Agent, Instrument, Result (winner)) = asymmetry (typical for meaningless morphemes)

Lexicon

1. place, where all irregularities are stored, where rules do not have access
2. morphemes are not predictable, must be defined, specified individually
3. Chomsky – especially in 1970s – Remarks on Nominalization; rejected DM as a field governed by rules
4. Aronoff – regularly formed words are not stored in the lexicon, all connected with relativized approach to productivity and blocking (what is regular cannot be blocked)
5. Rachel Lieber – introduced the notion of "permanent lexicon" – includes all unanalysable units (lexical terminal elements – final stage) – can be twofold – stems and affixes
6. emphasis on simplicity, units not complex
7. they are specified in the same way – several parameters:

1. phonological – cannot predict the form of the morpheme
2. semantic – cannot predict the meaning of a simple sign
3. stems – category (word class)

affixes – specified for subcategorization + specify the resulting category (verb)

4. syntactic feature – insertion frame – syntactic structure resulting from the use of suffix

run                This man runs a factory.

                             NP        V        NP

NP + gap

-ize insertion frame – He likes to generalize –NP + gap

He realized the circumstances – NP gap NP

5. diacritics – any specific information for run – run is of Anglo-Saxon origin

8. -ize = class 2 (no phonological change)
9. should be attached to lexical terminal elements

Halle

10. scheme of model of word-formation – 3 different lexicons:

1. list of morphemes
2. dictionary of words or complex words
3. constituted by means of the filter – indicates the gap in the system

Ellen

11. overgenerating principle

1. permanent lexicon – actual words are stored
2. conditional lexicon – includes all possible words (produced by regular WF rules)

12. recursiveness – implies infinite, list – implies finite = contradiction

Jackendoff

13. lexicon contains all the words – simple, complex, un/analysable, predictable, idiosyncratic

DiScullio & Williams

14. listemes – not words, not noticed by linguists before – also morphemes, affixes, longer units (idioms, collocations) are listemes
15. poses a problem – what is a lexical unit?

Bauer

16. detailed, concentrates on lexicon with idiosyncratic elements
17. phonological nature – segmental derivation (Monday – ei/i), suprasegmental feature (stress)
18. morphological nature
19. syntactic nature – adjectives may function as attributes (This is a nice house) or predicates (This house is nice)
20. former – only as attribute, afraid – only as predicate – should be captured in the lexicon
21. all these approaches take lexicon as a list

Steven Anderson

22. rejects the view of the lexicon as a list
23. we have to have knowledge – to be able to use lexicon

Qude

24. lexicon is the knowledge a speaker has of how words can instantiate positions in a syntactic structure
25. each word is constituted by its relations (paradigmatic and syntagmatic)
26. the meaning of a word is given by its relation to other words

David Pasetsky

27. rejects the existence of the lexicon, based on bracketing paradoxes (more unhappy/unhappy+er)
28. each lexical unit is described by its features on different levels

Onomasiological approach

        lexicon <-----> WF component

29. lexicon contains all morphemes, all affixes, all words produced by WF component
30. any idiosyncracy is realised in the lexicon

1. produce a brand new word
2. make use of existing word

31. transmission – produced by [V + ion] – the act, fact, degree of transmitting – shifted to the lexicon
32. car producers did not use new combination of morphemes – used an existing word and gave it a new meaning unpredictable from the rule (semantic shift)
33. clipping, acronyms produced in the lexicon (laboratory --> lab – remove a part of the form in lexicon)
34. hypothesis that only irregularly formed words are stored in the lexicon (idiosyncratic words + rules)
35. inflectional langs – words are stored with all their inflection or just their paradigm? all verbal conjugation forms? irregular forms of words, which are regular in WF? what is more economical? basic?
36. more advantageous to store just idiosyncracies and rules or to keep all the words with no subsequent operation?

LEVEL-ORDERING THEORIES

Level ordering

1. a theoretical principle explaining affix ordering, morpho-phoneme effects over affix boundaries and the varying effect of stress in morphologically complex words
2. Chomsky + Halle – The Sound Pattern of English (1968)
3. a phonological representation is seen as a single linear sequence of elements including also boundaries between words and morphemes
4. a morpheme boundary (symbolised +) = weak boundary
5. a word boundary (symbolised #) = strong boundary
6. in SPE, the distinction between the 2 kinds of boundary relies on phonological rather than on morphological evidence
7. do these phonologically motivated boundaries correspond with those morphologically motivated? – the work of Siegel, Allen and the lexical phonologists can be seen as an attempt to answer this question

Dorothy Siegel – Topics in Morphology (1979)

8. class 1 affixes (those attached with a + boundary, in SPE terms)
9. class 2 affixes (# boundary)
10. these two classes can be distinguished in terms of their phonological and morphological properties

class 1 – suffixes: +ion, +ity, +y, +al, +ic, +ate, +ous, +ive

                 prefixes: de+, re+, sub+, in+, con+, pre+, en+, be+

class 2 – suffixes: #ness, #less, #hood, #ful, #ly, #y, #like

                prefixes: de#, re#, sub#, un#, non#, semi#, anti#

11. the two classes of affixes give rise to different phonological effects
12. class 1 affixes carry a stress shift, class 2 are stress-neutral

class 1                                                       class 2

productive                productivity                 productiveness

13. class 1 affixes also tend to cause changes in the consonant and vowel segments of the stem to which they are attached

class 1                                            class 2 (do not trigger such changes)

fragile                        fragility                fragileness

14. morphological differences – class 1 affixes appear nearer to the root than class 2

Affix Ordering Generalization

15. class 1 affixes may attach to stems (i.e. bound morphemes)
16. class 2 affixes only ever attach to words

1 – re-fer                2 – re-take, un-fair

Siegel                 class 1 affixation

                                   lll

   stress rules

                                   lll

class 2 affixation

17. inflectional affixes in Eng. are class 2 – no stress change
18. the correctness of this prediction has been a matter of considerable debate

Allen (1978) – Extended Level Ordering Hypothesis

level 1 (+ affixation)

appropriate phonological rules including stress rules

level 2 (# affixation)

       level 3 (compounding)

     level 4 (regular inflection)

Kiparsky (1982) – Lexical Phonology & Morphology:

1. each morphological rule is accompanied by phonological rule – cyclic phonological rules at various levels (each level has a characteristic rule)

Lexical rules

19. apply only within words
20. cyclic – in each lvl it is necessary to pass through both lvls; output must be a legitimate word
21. structure-preserving
22. apply first
23. not automatic = conscious

Post-lexical rules

24. apply within words or across word boundaries (las(t) trip)
25. not cyclic
26. not necessarily structure-preserving
27. apply later
28. automatic – whenever conditions are met

Elsewhere condition (Kiparsky)

29. the more specific rule applying at a lower level blocks the application of more general rule operating at a higher level
30. this explains why plural foot – feet is assigned at level 1 where irregular plural (specific rule) is formed and regular plural formation *foots at level 3 is blocked
31. the failure of blocking results in doublets like dreamt/dreamed, syllabi/syllabuses
32. the bracketing paradox – un – [ grammatical – ity ] = incorrect

        [ un – grammatical ] – ity = correct (un attaches to Adj only)

NATURAL MORPHOLOGY

1. introduced in 1980-90s
2. compared to other approaches – different perspective to linguistic issues
3. natural = cognitively simple, elementary, accessible – universally preferred
4. if it is simple or not – matter of scale
5. 3 filters for naturalness:

1. universal filter – universal rules

6. condition the use of language
7. cognitive foundations of language and its use, psychological conditioning of using lang, sociolinguistic factors, semiotic motivation
8. internal and external factors are equally important

2. linguistic factors outside morphology
3. morphological system of the language itself

9. NM drew on former approaches – markedness theory

boy – unmarked = natural                boys – marked = unnatural (plural marked)

10. unmarked categories are more natural and frequent in langs

1. preference for unmarked categs – they are more easily acquired

11. children regularize irregularities – according to boy/boys (unmarked) they form *oxes (instead of marked oxen)

12. numbers of preferences are related to semiotics – Peirce's classification (see Essentials of English Linguistics)
13. icons = signs based on similarities between the word and the object
14. crucial role in NM – images (most iconic icons), diagrams, metaphors

Images

15. Berko (psycholinguist) – non-existing word wug – respondents of various ages had to create diminutives
16. children --> wig – relation between sound and meaning (i = front hight vowel indicating smallness)
17. older --> wuggie – u preserved, i sound added (smallness)
18. oldest --> wuglet – e sound still front but not high
19. diagrams – signs based on analogy of relation between signans (form) and signatum (meaning)

(semantically) stone wall – stone = modifier, wall = head

formal PoV – stone = modifier, wall = head

--> there is an analogy of relation between modifier and head --> principle of diagramaticity

2. preference for constructional iconicity

teach --> teach-er = new form (by suffix) represents new meaning (natural)

tooth --> teeth = new meaning by internal modification (less natural)

to cut --> a cut = new meaning but no new form (unnatural)

backformation --> to backform = new meaning by deleting form (unnatural)

--> most natural processes = new meaning represented by adding some new form

Indexes

20. direct relation between signans and signatum

teach-er = direct attaching

abso-fucking-lutely – word absolutely (base) violated – new meaning inserted

3. preference for indexicality

21. fixed order of morphemes – cannot add affixes arbitrarily

4. preference for morphosemantic transparency

22. we can identify the semantics from the constituent morphemes

boy-s – meaning of the whole = meaning of its components together – Freje – principle of compositionality

        the + cat + sat + on + the + mat – meaning of the words --> meaning of the sentence

        high school --> special type of school? (compound)

                         --> high building? – compositionality (high + school)

5. preference for morphosyntactic transparency and opacity

degrees:

1. both modifier and head are morphosyntactically transparent (doorbell – both preserve original meaning) = natural
2. modifier is not MST, head is (strawberry) = less natural
3. modifier is MST, head is not (jail-bird – a person) = less natural
4. none of them is MST (ladybird – kind of insect) = unnatural

6)         preference for morphotactic transparency

23. it is natural to have complex word, in which the form of the motivating constituent is not changed

teach – teacher – from the formal PoV – no change in teach or er

roast – roaster – syllabic border has moved (roast --> roas-ter) = resyllabification

black + board – separately – stress on both

                 --  compound – change of stress (this is against MTT)

divide – divisible – change of diphthong into monophthong, consonantal change = MTT much lower = unnatural

suppletion (bad/worse, be/am/is/are) – formal relation is non-existent = unnatural

7)         preference for biuniqueness

24. 1 form should represent 1 meaning (-able = only Quality)
25. -er = Agent/Instrument/Location (eg Londoner) – 1 form with more meanings
26. Agent = er/ist/ian/ee/man... – 1 meaning by more forms
27. this is unnatural (many relations)

MORPHOLOGICAL TYPOLOGY

1. deals with comparisons of various languages (langs with rich M (Slavonic), langs with hardly any M, langs with many/no word classes...)
2. study of similarities and differences of langs
3. results in certain classification
4. related to language universals – features shared by all/majority of natural langs
5. no strict boundary between MT & LU (just quantitative)
6. number of langs in particular type is smaller (?) than in LU
7. cognitive abilities of speakers are basically the same
8. Chomsky – we have precondition for lang (tabula rasa)
9. there is no lang without phonemes (divided into vowels and consonants), morphemes, sentences...
10. secondary categories, word order, other individual grammatical phenomena – different
11. 19th C – period of exclusive focus on comparison of langs
12. 20th C – significant results – thanks to American linguists – mapping the lang (USA – native lang) – trying to preserve langs dying out
13. present days – thousands of langs die (ppl have to adapt to majority – quit speaking their lang)
14. other accounts of LU – monogenetic theory (of protolanguage) – not justified
15. other theory – existence of LU follows from functional & pragmatic function of lang – rejected (synonyms, homonyms – not pragmatic)
16. language is not pragmatic – many redundancies, ambiguity...
17. Chomsky – deductive method of reaching LU – study 1 lang in detail & draw generalizations (theoretical approach to LU)
18. Greenberg – study many langs & compare them by parameter & then generalize
19. it is hard to identify a well-balanced lang sample – usually more than 50 langs for a statistically relevant research

1. absolute U – shared by all the langs

2. statistical U – prevailing number of langs have certain features (almost all langs have  

        suffixation – suffixation is statistically universal)

3. implicational U – if A --> then B (if there is feature A, then there has to be feature B)

20. at least 2 colour terms in each lang (black and white); 3 colours (black white red); no lang with 4 colour terms; 5 (B W R yellow green)

black                                yellow          --> if there is a word for yellow, then there must be words for B W R too

                red                                (they are before it)

        white                                green

21. if a lang has discontinuous affixes – always has prefixes or suffixes (or both) = natural
22. unnatural = eg German – split affixes (ge-arbeit-et)
23. if a lang has D, it always has I

Morphological Classification

Friedrich Schlegel – 1st to propose relevant MC (organic & agglutinative langs) – 19th C

24. organic – ideal, the substance of lang (lexical meaning & its formal representation) are in harmony with the spirit of the lang (grammatical meaning & its formal representation)

leipein ("let")

1st sg – leipo                        --> any new grammatical meaning is expressed twice – in

past – el ip on                      substance (stem) and in suffix

perfect – la loip a

25. agglutinative – combines several affixes in succession, each of them has only one meaning --> long words
26. eg Hungarian, Turkish,...

Edward Sapir (American descriptivism) – analytic, synthetic (inflectional and agglutinative),

polysynthetic, introflecting langs

Analytic (isolating) languages

27. most typical (English, Chinese, Vietnamese...)
28. do not have much I (ideal langs – no I)
29. cases expressed by prepositional phrases (no I morphemes) – to my house/from my house
30. conjugation of V requires use of personal pronouns (I read, you read)
31. infinitive expressed by separate words (to read)
32. word order – WO expresses relations between words (because it is impossible to express them with words themselves)

Agglutinative (synthetic) languages

33. rich in affixes (mostly ideal signs – 1 form = 1 meaning)

er    –   ler  –  im    –    iz         –       de

               house   plural   my   plural of my   local case        --> in our house (Turkish)

34. no grammatical homonymy/synonymy/polysemy
35. unlike Slavonic langs – eg chlapa – several meanings (M, animate, sg, A/G case (gram. polysemy)); Agent – ee/er/ian/ist/... --> gramm. synonymy

Polysynthetic languages

36. American Indian, Australian
37. combines many lexical D & I morphemes in one word
38. can express by 1 word what is expressed by sentence in other langs (very long words – 10-15 affixes)

Introflecting languages

39. based on internal flection (no affixes – only changes in stem)
40. Arabic, Hebrew – vowels inserted into consonantal skeleton