Faithfulness and Markedness

1
Faithfulness and Markedness

• The Plan for Today
• Markedness and Faithfulness and conflict
  resolution in OT
• Two factorial typologies in phonology
• - syllable structure
• - syllable peaks and margins

2
Faithfulness and Markedness

• Presentation of Grimshaws account of do-support
  in English, as a result of the interaction of
  faithfulness and markedness
• Presentation of Pesetskys deletion theory
• A reflection on markedness hierarchies in syntax

3
Markedness in Phonology

• A marked segment is less natural, more complex,
  less common, not basic, not stable, it appears
  later in language acquisition, in fewer
  languages, in less positions, it is subject to
  neutralization, harder to articulate,
  perceptually more salient, etc.

4
Examples of markedness hierarchies

• Sonority hierarchy
• Nasal vowels are more marked than oral ones
• Voiced obstruents are more marked than voiceless
  ones
• Voiced sonorants are more marked than voiceless
  ones

5
Some diagnostics for markedness

• - Implication the marked feature implies the
  unmarked one.
• - Frequency unmarked features are more frequent
  than marked ones.
• - Processes the unmarked features emerge only
  under special circumstances (otherwise emergence
  of the unmarked), the unmarked feature
  neutralizes more easily.

6
Markedness vs. Faithfulness

• Since markedness requires simple structures and
  faithfulness requires that inputs always be
  realized the way they come, regardless of their
  complexity, the two kinds of constraints often
  impose conflicting demands on outputs.
• Depending on which win, structures are faithful
  or unmarked.

7
Markedness vs. Faithfulness

• OT is especially good at the resolution of
  conflicts between markedness and faithfulness.
• Different languages make different choices, and
  the possible solutions are accounted for by the
  same constraints but ranked differently.
• Ideally all possible rankings define possible
  grammars and all possible grammars are the
  results of different rankings.

8
Simple factorial typology 1

• Syllables without onsets and with codas are
  tolerated in many languages, but only if there is
  no other way to syllabify. A sequence VCV is in
  the default case (and in all languages)
  syllabified as V.CV.
• (Jakobsons typology) all languages have CV
  syllables.
• French, German and English tolerate syllables
  without onsets and with codas , but only if there
  is no way out, that is if no other
  syllabification is better, because of higher
  ranking constraints (faithfulness).

9
Simple factorial typology 1

• Strategies to obtain a better (less marked
  syllable structure). Remember hiatus avoidance
• Monomorphemic VCV are always syllabified as V.CV
  or VCambysyllV but never as VC.V
• E coma, comma
• G Koma, Komma
• F coma (no ambisyllabicity)

10
Simple factorial typology 1

• Some languages do not have syllables without
  onsets, or ones with codas, and systematically
  use strategies to enhance bad syllables, for
  instance in loanwords.
• Japanese kurimasu for Christmas and arubatu for
  Arbeit.
• How does OT account for these observations?

11
Simple factorial typology 1

• In considering hiauts avoidance, we have seen
  that nearly no language applies pure faithfulness
  or pure markedness, but that finer distinctions
  must be made which consider morphological and
  phonological facts.

12
Simple factorial typology 1

• The factorial typology shown here is thus an
  idealization.
• The constraint ranking necessary to account for
  hiatus is shown to be more general since it also
  applies to other segment strings without hiatus.

13
Simple factorial typology 1

• Prince Smolensky (1993) on syllabification
• Markedness Constraints ONSET, NOCODA,
• Faithfulness Constraints MAX, DEP
• Through permutation of these constraints,
  different kinds of syllables are obtained.

14
Faithfulness (/V/)
15
Epenthesis (/V/)
16
Deletion (/V/)
17
Factorial typology as constraint reranking

• With 3 constraints, 6 permutations are possible,
  but only 3 are distinct. The other ones deliver
  the same results.
• Given 4 constraints ONSET, NOCODA, MAX, and DEP,
  what are the possible grammars?
• 24 (4!) possible permutations of them. But not
  all 24 are interesting, because some constraints
  never conflict with each other.

18
Factorial Typology as constraint reranking

• ONSET and NOCODA, for instance, do not conflict
  (they both impose .CV.)
• If both MAX and DEP dominate ONSET and NOCODA,
  the ordering of the faithfulness constraints is
  indifferent.
• MAX gtgt DEP gtgt ONSET , NOCODA and DEP gtgt MAX
  gtgt ONSET , NOCODA are equivalent.
• MAX conflicts with DEP only if Onset or NOCODA
  dominates one of them.

19
Deletion (/tai/)

• /tai/ Onset/NoCoda DEP MAX
• ________________________________________
• tai !
• tati !
• gt ta

20
Epenthesis (/tai/)

• /tai/ MAX Onset/NoCoda DEP
• ________________________________________
• tai !
• gt tati
• ta

21
Faithfulness (/tai/)

• /tai/ MAX DEP Onset/NoCoda
• ________________________________________
• gt tai
• tati !
• ta !

22
Deletion (/takti/)

• /takti/ Onset/NoCoda DEP MAX
• ________________________________________
• takti !
• takati !
• gt tati

23
Epenthesis (/takti/)

• /takti/ MAX Onset/NoCoda
  DEP
• ________________________________________
• takti !
• gt takati
• tati !

24
Faithfulness (/takti/)

• /takti/ MAX DEP Onset/NoCoda
• ________________________________________
• gt takti
• takati !
• tati !

25
Factorial Typology and possible grammars

• Notice that, given the set of constraints, no
  language can exist that has only syllables with
  codas but without onsets. This has nothing to do
  with the orderings, but with the kinds of
  constraints used.

26
Factorial Typology and possible grammars

• Given the fact that all languages have Cs and
  Vs, there is also no language without CV
  syllables.
• And VCV will aways be syllabified as V.CV, as
  desired.

27
Simple factorial typology 2

• Second factorial typology Syllable peaks (PS)
• Syllable peaks are the most sonorous part of the
  syllables
• s or
  s
• / \ / \
• C V C C C (C)
• Which kind of segments may be C?

28
Simple factorial typology 2

• French has only vowels as syllable peaks words
  like ocre or siècle have a coda with increasing
  sonority.
• Though such codas are marked and impossible in
  most languages, this is still better than make
  the liquid syllabic.

29
Simple factorial typology 2

• German and English have vowels and sonorants as
  syllable peaks
• G Himmel, eben, nieder
• E eagle, meter, button
• Imdlawn Tashlhiyt Berber tolerates all segments
  as syllable peaks, also obstruents

30
Imdlawn Tashlhiyt Berber Data from Dell
Elmedlaoui (1985)

• Voiceless stop .rA.tK.tI. ra-t-kti
• Voiced stop .bD.dL. bddl
• Voiceless fricative .tF.tKt. t-ftk-t
• .tX.zNt. t-xzn-t
• Voiced fricative .txZ.nAkkw. t-xznnakkw
• ratkti she will remember, bddl exchange,
  .tftkt. you suffered a strain, txznt you
  stored, txZ.nAkkw she even stockpiled

31
Imdlawn Tashlhiyt Berber

• Nasal .tzMt. t-zmt
• Liquid .tR.gLt. t-rgl-t
• High vowel .Il.dI. i-ldi
• .rat.lUlt. ra-t-lul-t
• Low vowel .tR.bA. t-rba
• tzmt it (f.) is stifling, trglt you
  locked, ildi he pulled, ratlult yoou will be
  born, trba she carried on her back

32
Imdlawn Tashlhiyt Berber

• Domain for syllabification is the Phonological
  Phrase
• All syllables have an onset, except for the
  phrase initial ones, which can be onsetless. This
  is accounted for by an alignment constraint which
  is not important here.

33
Simple factorial typology 2 Syllable peaks

• Universal margin and peak hierarchies
• M/a gtgt M/i gtgt M/l gtgt M/t
• P/t gtgt gtgt P/l gtgt P/i gtgt P/a
• This kind of nonpermutable hierachies
  (meta-constraints, as McCarthy calls them) are
  different from the usual hierarchies. They are
  defined by extragrammatical factors, here the
  acoustic properties of individual segments.

34
Simple factorial typology 2 Syllable peaks

• As a consequence, no language allows obstruents
  as syllable peaks, but not vowels, for instance.
• Languages will chose a place in these
  hierarchies, from which on no segment can be peak
  or margin

35
Simple factorial typology 2 Syllable peaks

• Untenable margins
• M/a gtgt MAX gtgt M/i gtgt M/l gtgt M/t
• The maximum sonority of possible onsets/codas
  cannot be higher than the sonority level i.
  (Berber)
• M/a gtgt M/i gtgt MAX gtgt M/l gtgt M/t
• The maximum sonority of possible onsets/codas
  cannot be higher than the sonority level l.

36
Simple factorial typology 2 Syllable peaks

• Untenable peaks
• MAX gtgt P/t gtgt P/n gtgt P/l gtgt P/i gtgt P/a
• The minimum sonority of possible nuclei cannot be
  less than the sonority level of t. (Berber)
• P/t gtgt P/n gtgt P/l gtgt MAX gtgt P/i gtgt P/a
• The minimum sonority of possible nuclei cannot be
  less than the sonority level of i. (French,
  German, English)

37
Factorial typology in phonologySummary

• OT provides a good framework to show how
  conflicts are resolved, to show what possible
  grammars are and to show that unmarked structures
  are the aim of every language, even if, in some
  circumstances, the aim of unarkedness is
  concealed by higher requirements of faithfulness
  (Emergence of the Unmarked).

38
Faithfulness and Markedness in syntax

• A first example
• Grimshaws account of English basic clause
  structure
• Goal
• Explanation of the basic patterns of English
  clause structure
• When does do-support apply? When can faithfulness
  be violated?

39
The data to be explained

• John saw Mary No do-support
• John did see Mary in simple clauses
• John saw not Mary Do support in
• John did not see Mary negative clauses
• who came? No do-support in
• who did come? Subject questions
• who likes he? Do-support in
• who does he like? Object questions

40
Simple assertions

• Input John, left
• or, rather, the PAS corresponding to it.
• GEN
• Merge, Move, Insert functional categories

41
Simple assertions

• Some candidates
• VP John left
• IP Infl VP John left
• IP John Infl VP t left
• IP John left VP t t
• IP John did VP t leave
• IP Did VP John leave
• IP Left VP John t

42
Obligatory Heads

• Recall that the input consists of PAS only. The
  addition of functional layers above VP is thus a
  faithfulness violation.
• OblHd A projection must have a visible head

43
A First Tableau

44
A Second Principle

• OblHd A projection must have a visible head
• Full Interpretation Do not insert meaningless
  elements/PAS must be respected

45
A Second Tableau

46
A Third Principle

• OblHd A projection must have a visible head
• Full Interpretation Do not insert meaningless
  elements/PAS must be respected
• Econ/Stay! Trace is ungrammatical/Do not move!

47
A Third Principle

• Econ/Stay may be considered a faithfulness
  constraint, too, because traces are not in the
  input.
• On the other hand, it may also be profitable to
  view it as a markedness constraint (for
  derivations)

48
A Third Tableau

49
The kernel sentence

• When no markedness constraints are applicable,
  the sentence cannot and does not project beyond
  VP (beyond the input).
• Note that this is special kind of economy (do not
  project beyond what is absolutely necessary),
  which is not represented in Minimalism.
• There is a possible processing argument for
  OblHd.

50
Simple Negation

• Input
• John, not, left
• Some candidates
• not John left
• NegP John not t left
• IP John left not t t
• IP John Infl not t left
• IP John did not t leave
• IP Did not John leave

51
Negation the Principles Introduced so far

52
New Principles

• The faithfulness principles considered so far do
  not yield the correct result.
• We need to consider a further principle, a
  markedness constraint
• Subject (vulgo EPP)
• The highest A-specifier of a sentence must be
  filled!

53
Negation the New Principle

54
New Principles

• SUBJ already implies that the subject is the
  highest element in the clause --but not any
  structural position will do.
• Case, a second markedness principle,
• restricts options to Spec,VP and Spec,IP
• Case (vulgo Case Filter)
• The highest position of a chain must have Case!

55
Negation the New Principles

56
New Principles

• Case (vulgo Case Filter)
• The highest position of a chain must have Case!
• LexEconLexical Trace!

57
Negation the New Principles

58
Brief summary

• Faithfulness
• OblHd, Full Interpretation, Econ, LE
• Note the relation between Econ and LE
• Perhaps
• Constraint patterns which can be specified to
  different degrees
• Markedness
• Subject, Case

59
An attempt of a factorial typology 1

• LE gt FI
• No movement of the verb, expletive insertion
• John did not kiss Mary
• FI gt LE
• No expletives but verb movement
• Jean nembrasse pas Marie

60
An attempt of a factorial typology

• Econ gt OblHd
• Neither expletives nor movement
• ---- COMP Jan inte har drukket kaffee
• John not has drunk coffee

61
An attempt of a factorial typology

• LE gt OblHd gt FI English
• LE gt FI gt OblHd Danish
• FI gt LE gt OblHd Danish
• FI gt Oblhd gt LE French
• OblHd gt LE gt FI English
• OblHd gt FI gt LE French

62
An attempt of a factorial typology

• This reasoning presupposes that we have an IP,
  and not a NegP
• If Case is Low ...

63
More data wh-questions

• OpSpec Wh-phrases c-command the extended
  V-projection they modify
• OpSpec Econ
• wh-in situ
• wh moved

64
In Situ Languages
65
Do-Insertion in wh-questions

• We now want to check if the constraints
  considered so far make correct predictions
  concerning do-support in wh-questions.
• No do-support in subject questions
• Do-support in object questions, etc.

66
Subject Questions
67
Object Questions
68
Further faithfulness considerations

• FI and Econ disfavor the insertion of elements
  which are not in the input
• There should also be principles that require that
  input material be realized.
• These are principles of the Parse or MAX-family
• We will study them in the light of Pesetskys
  economy of pronunciation approach

69
Placement of functional heads

• Before we can consider deletion, we should
  identify a further markedness principle
• Complement clauses are introduced by a
  complementizer
•  
• Je crois que Pierre a faim
• Ich denke dass Peter Hunger hat
• I think that Peter is hungry

70
Placement of functional heads

• PPs are introduced by a preposition
• avec Pierre
• mit Peter
• with Peter

71
Placement of functional heads

• DPs are introduced by a determiner
• le livre
• das Buch
• the book

72
LEFT!

• The Generalization
•  
• Left Edge of Functional (Sub-)Trees (LEFT)
• Projections of functional categories are
  introduced by their heads

73
Placement of functional heads

• Left edges of CPs
•  
• Je crois que Pierre a faim
• I think that Pierre has hunger
• Je crois Pierre a faim
•  
• Ich denke dass Peter Hunger hat
• I think that Peter hunger has
• Ich denke Peter Hunger hat

74
Placement of functional heads

75
Placement of functional heads

• Relative clauses
•  lhomme qui je connais
• the man who I know
• lhomme que je connais
• the man that I know

76
Relative clauses

77
Rec disfavors deletion

• BUT
• l'homme avec qui j'ai dansé
• l'homme que j'ai dansé
•  
• REC (Pesetsky)
• A unit with semantic content must be pronounced
  unless it has a sufficiently local antecedent

78
Some Optimality Effects in Syntax

• REC belongs to the family of faithfulness
  constraints
• Rec gt Left gt Tel
• What is recoverable and what not is dealt with
  differently in different languages

79
No deletion with PP-relatives

80
Deletion with NP-relatives

81
More Faithfulness

• Parse (wh)
• wh-features of the input must be realized in the
  output
• ParseScope
• The scope of a wh-phrase (etc.) must correspond
  to the input
• Parse (Case)
• E.g. lexical exceptions must not be changed

82
More Markedness

• Scope
• If X c-commands Y, then X takes scope over Y
• Gov
• Traces must be governed
• Agree!
• The verb agrees with the nominative DP

83
Markedness Hierarchies

• CASE
• nominative lt accusative lt dative
• See regular Case frames in German
• nom, nom acc, nom acc dat
• AGREEMENT
• nominative lt accusative lt dative
• GERMAN - HUNGARIAN - BASQUE

84
Markedness Hierarchies

• ANAPHORIC BINDING
• nominative lt accusative lt dative
• dass er den Frauen einander zeigt
• dass er die Frauen einander zeigt
• that he the women each other shows
• MOVEMENT
• (The Keenan-Comrie Hierarchy)
• nom lt acc lt dat

85
Markedness Hierarchies

• WORD ORDER (among pronouns)
• nominative lt accusative lt dative
• Minimal Link Effects?
• nom lt acc
• who saw what
• what did who see?

86
Markedness Hierarchies

• Markedness Hierarchy for choice of antecedent for
  anaphors
• No person lt 3.ps lt 1./2. ps
• soi-même -- sich - sie
• Markedness Hierarchy for choice of antecedent for
  anaphors
• Clausemate lt ECM-infinitive lt control infinitive
• English - German - Danish

87
Markedness Hierarchies

• Markedness Hierarchy for choice of landing site
  for wh- movement
• Clausemate lt Tensed infinitive lt subjunctive lt
  indicative
• Warlpiri - Russian X - Russian Y -English