derivations - massachusetts institute of...
TRANSCRIPT
![Page 1: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/1.jpg)
Derivations
Paul Kiparsky
![Page 2: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/2.jpg)
Argue withme!
![Page 3: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/3.jpg)
And witheach other!
![Page 4: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/4.jpg)
Derivations in phonology: a brief history
1 Pan. ini (ca. 500 B.C.), Chomsky (1951), Halle (1962):ordered rules derive phonetic representations fromunderlying representations.
2 Stanley (1967): ordered rules plus morpheme structureconstraints (well-formedness conditions).
Argument 1: rules impose an arbitrary direction ofdependence on co-occurrence restrictions.Argument 2: rule ordering is not needed for phonotactics.
3 Kisseberth (1970): proposed to solve the CONSPIRACIES
PROBLEM by OUTPUT CONSTRAINTS, allowing the sharedcontext of rules be factored out.
![Page 5: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/5.jpg)
Derivations in phonology: a brief history
1 Pan. ini (ca. 500 B.C.), Chomsky (1951), Halle (1962):ordered rules derive phonetic representations fromunderlying representations.
2 Stanley (1967): ordered rules plus morpheme structureconstraints (well-formedness conditions).
Argument 1: rules impose an arbitrary direction ofdependence on co-occurrence restrictions.Argument 2: rule ordering is not needed for phonotactics.
3 Kisseberth (1970): proposed to solve the CONSPIRACIES
PROBLEM by OUTPUT CONSTRAINTS, allowing the sharedcontext of rules be factored out.
![Page 6: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/6.jpg)
Derivations in phonology: a brief history
1 Pan. ini (ca. 500 B.C.), Chomsky (1951), Halle (1962):ordered rules derive phonetic representations fromunderlying representations.
2 Stanley (1967): ordered rules plus morpheme structureconstraints (well-formedness conditions).
Argument 1: rules impose an arbitrary direction ofdependence on co-occurrence restrictions.Argument 2: rule ordering is not needed for phonotactics.
3 Kisseberth (1970): proposed to solve the CONSPIRACIES
PROBLEM by OUTPUT CONSTRAINTS, allowing the sharedcontext of rules be factored out.
![Page 7: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/7.jpg)
Derivations in phonology: a brief history
1 Pan. ini (ca. 500 B.C.), Chomsky (1951), Halle (1962):ordered rules derive phonetic representations fromunderlying representations.
2 Stanley (1967): ordered rules plus morpheme structureconstraints (well-formedness conditions).
Argument 1: rules impose an arbitrary direction ofdependence on co-occurrence restrictions.Argument 2: rule ordering is not needed for phonotactics.
3 Kisseberth (1970): proposed to solve the CONSPIRACIES
PROBLEM by OUTPUT CONSTRAINTS, allowing the sharedcontext of rules be factored out.
![Page 8: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/8.jpg)
Derivations in phonology: a brief history
1 Pan. ini (ca. 500 B.C.), Chomsky (1951), Halle (1962):ordered rules derive phonetic representations fromunderlying representations.
2 Stanley (1967): ordered rules plus morpheme structureconstraints (well-formedness conditions).
Argument 1: rules impose an arbitrary direction ofdependence on co-occurrence restrictions.Argument 2: rule ordering is not needed for phonotactics.
3 Kisseberth (1970): proposed to solve the CONSPIRACIES
PROBLEM by OUTPUT CONSTRAINTS, allowing the sharedcontext of rules be factored out.
![Page 9: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/9.jpg)
Derivations in phonology: a brief history
1 Kisseberth’s translation of “functional unity” into formalsimplicity was only partially successful, because
you can’t simplify rules which are triggered by a constraint,rules can create prohibited configurations if the output isrepaired by a subsequent rule, andthere was no substantive theory of targets.
2 Stampe (1972/1979): back to a strictly processual account,even of phonotactics. Distinguish RULES (learned) andPROCESSES (innate). Conflicts between them resolved bylimitation, suppression, and ordering.
3 Constraint-driven serial derivations: constraints and repairprocesses (Paradis 1987, 1988), Calabrese 1995,Harmonic Phonology (Goldsmith 1991), HarmonicSerialism (McCarthy 2007).
![Page 10: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/10.jpg)
Derivations in phonology: a brief history
1 Kisseberth’s translation of “functional unity” into formalsimplicity was only partially successful, because
you can’t simplify rules which are triggered by a constraint,rules can create prohibited configurations if the output isrepaired by a subsequent rule, andthere was no substantive theory of targets.
2 Stampe (1972/1979): back to a strictly processual account,even of phonotactics. Distinguish RULES (learned) andPROCESSES (innate). Conflicts between them resolved bylimitation, suppression, and ordering.
3 Constraint-driven serial derivations: constraints and repairprocesses (Paradis 1987, 1988), Calabrese 1995,Harmonic Phonology (Goldsmith 1991), HarmonicSerialism (McCarthy 2007).
![Page 11: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/11.jpg)
Derivations in phonology: a brief history
1 Kisseberth’s translation of “functional unity” into formalsimplicity was only partially successful, because
you can’t simplify rules which are triggered by a constraint,rules can create prohibited configurations if the output isrepaired by a subsequent rule, andthere was no substantive theory of targets.
2 Stampe (1972/1979): back to a strictly processual account,even of phonotactics. Distinguish RULES (learned) andPROCESSES (innate). Conflicts between them resolved bylimitation, suppression, and ordering.
3 Constraint-driven serial derivations: constraints and repairprocesses (Paradis 1987, 1988), Calabrese 1995,Harmonic Phonology (Goldsmith 1991), HarmonicSerialism (McCarthy 2007).
![Page 12: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/12.jpg)
Derivations in phonology: a brief history
1 Kisseberth’s translation of “functional unity” into formalsimplicity was only partially successful, because
you can’t simplify rules which are triggered by a constraint,rules can create prohibited configurations if the output isrepaired by a subsequent rule, andthere was no substantive theory of targets.
2 Stampe (1972/1979): back to a strictly processual account,even of phonotactics. Distinguish RULES (learned) andPROCESSES (innate). Conflicts between them resolved bylimitation, suppression, and ordering.
3 Constraint-driven serial derivations: constraints and repairprocesses (Paradis 1987, 1988), Calabrese 1995,Harmonic Phonology (Goldsmith 1991), HarmonicSerialism (McCarthy 2007).
![Page 13: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/13.jpg)
Derivations in phonology: a brief history
1 Kisseberth’s translation of “functional unity” into formalsimplicity was only partially successful, because
you can’t simplify rules which are triggered by a constraint,rules can create prohibited configurations if the output isrepaired by a subsequent rule, andthere was no substantive theory of targets.
2 Stampe (1972/1979): back to a strictly processual account,even of phonotactics. Distinguish RULES (learned) andPROCESSES (innate). Conflicts between them resolved bylimitation, suppression, and ordering.
3 Constraint-driven serial derivations: constraints and repairprocesses (Paradis 1987, 1988), Calabrese 1995,Harmonic Phonology (Goldsmith 1991), HarmonicSerialism (McCarthy 2007).
![Page 14: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/14.jpg)
Derivations in phonology: a brief history
1 Kisseberth’s translation of “functional unity” into formalsimplicity was only partially successful, because
you can’t simplify rules which are triggered by a constraint,rules can create prohibited configurations if the output isrepaired by a subsequent rule, andthere was no substantive theory of targets.
2 Stampe (1972/1979): back to a strictly processual account,even of phonotactics. Distinguish RULES (learned) andPROCESSES (innate). Conflicts between them resolved bylimitation, suppression, and ordering.
3 Constraint-driven serial derivations: constraints and repairprocesses (Paradis 1987, 1988), Calabrese 1995,Harmonic Phonology (Goldsmith 1991), HarmonicSerialism (McCarthy 2007).
![Page 15: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/15.jpg)
Optimality Theory
1 Prince & Smolensky (1993): ranked constraints uniquelydetermine the processes that implement them.
2 Constraints are violable, but violation is minimal.
3 Constraints can both “trigger” and “block” processes.Schematically:
P → Q is triggered in the context X___Y if *XPY ≫ *Q,P → Q is blocked in the context X___Y if *XQY ≫ *P.
4 The constraint system evaluates output representations.(Under Harmonic Serialism, it evaluates each step in aderivation.)
5 A ranking determines a grammar, the possible rankingsdetermine the typological space.
6 To the extent that constraints are grounded and universal,OT is a theory of naturalness.
![Page 16: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/16.jpg)
Optimality Theory
1 Prince & Smolensky (1993): ranked constraints uniquelydetermine the processes that implement them.
2 Constraints are violable, but violation is minimal.
3 Constraints can both “trigger” and “block” processes.Schematically:
P → Q is triggered in the context X___Y if *XPY ≫ *Q,P → Q is blocked in the context X___Y if *XQY ≫ *P.
4 The constraint system evaluates output representations.(Under Harmonic Serialism, it evaluates each step in aderivation.)
5 A ranking determines a grammar, the possible rankingsdetermine the typological space.
6 To the extent that constraints are grounded and universal,OT is a theory of naturalness.
![Page 17: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/17.jpg)
Optimality Theory
1 Prince & Smolensky (1993): ranked constraints uniquelydetermine the processes that implement them.
2 Constraints are violable, but violation is minimal.
3 Constraints can both “trigger” and “block” processes.Schematically:
P → Q is triggered in the context X___Y if *XPY ≫ *Q,P → Q is blocked in the context X___Y if *XQY ≫ *P.
4 The constraint system evaluates output representations.(Under Harmonic Serialism, it evaluates each step in aderivation.)
5 A ranking determines a grammar, the possible rankingsdetermine the typological space.
6 To the extent that constraints are grounded and universal,OT is a theory of naturalness.
![Page 18: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/18.jpg)
Optimality Theory
1 Prince & Smolensky (1993): ranked constraints uniquelydetermine the processes that implement them.
2 Constraints are violable, but violation is minimal.
3 Constraints can both “trigger” and “block” processes.Schematically:
P → Q is triggered in the context X___Y if *XPY ≫ *Q,P → Q is blocked in the context X___Y if *XQY ≫ *P.
4 The constraint system evaluates output representations.(Under Harmonic Serialism, it evaluates each step in aderivation.)
5 A ranking determines a grammar, the possible rankingsdetermine the typological space.
6 To the extent that constraints are grounded and universal,OT is a theory of naturalness.
![Page 19: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/19.jpg)
Optimality Theory
1 Prince & Smolensky (1993): ranked constraints uniquelydetermine the processes that implement them.
2 Constraints are violable, but violation is minimal.
3 Constraints can both “trigger” and “block” processes.Schematically:
P → Q is triggered in the context X___Y if *XPY ≫ *Q,P → Q is blocked in the context X___Y if *XQY ≫ *P.
4 The constraint system evaluates output representations.(Under Harmonic Serialism, it evaluates each step in aderivation.)
5 A ranking determines a grammar, the possible rankingsdetermine the typological space.
6 To the extent that constraints are grounded and universal,OT is a theory of naturalness.
![Page 20: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/20.jpg)
Optimality Theory
1 Prince & Smolensky (1993): ranked constraints uniquelydetermine the processes that implement them.
2 Constraints are violable, but violation is minimal.
3 Constraints can both “trigger” and “block” processes.Schematically:
P → Q is triggered in the context X___Y if *XPY ≫ *Q,P → Q is blocked in the context X___Y if *XQY ≫ *P.
4 The constraint system evaluates output representations.(Under Harmonic Serialism, it evaluates each step in aderivation.)
5 A ranking determines a grammar, the possible rankingsdetermine the typological space.
6 To the extent that constraints are grounded and universal,OT is a theory of naturalness.
![Page 21: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/21.jpg)
Optimality Theory
1 Prince & Smolensky (1993): ranked constraints uniquelydetermine the processes that implement them.
2 Constraints are violable, but violation is minimal.
3 Constraints can both “trigger” and “block” processes.Schematically:
P → Q is triggered in the context X___Y if *XPY ≫ *Q,P → Q is blocked in the context X___Y if *XQY ≫ *P.
4 The constraint system evaluates output representations.(Under Harmonic Serialism, it evaluates each step in aderivation.)
5 A ranking determines a grammar, the possible rankingsdetermine the typological space.
6 To the extent that constraints are grounded and universal,OT is a theory of naturalness.
![Page 22: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/22.jpg)
Optimality Theory
1 Prince & Smolensky (1993): ranked constraints uniquelydetermine the processes that implement them.
2 Constraints are violable, but violation is minimal.
3 Constraints can both “trigger” and “block” processes.Schematically:
P → Q is triggered in the context X___Y if *XPY ≫ *Q,P → Q is blocked in the context X___Y if *XQY ≫ *P.
4 The constraint system evaluates output representations.(Under Harmonic Serialism, it evaluates each step in aderivation.)
5 A ranking determines a grammar, the possible rankingsdetermine the typological space.
6 To the extent that constraints are grounded and universal,OT is a theory of naturalness.
![Page 23: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/23.jpg)
The “derivational residue”
1 Two major problemsOpacityCyclicity
2 Two types of solutions
Introduce transderivational faithfulness constraints(Sympathy, O/O) and/or transderivational constraintsformulated over faithfulness relations (OT-CC).Modularity: level-ordered cascade of classic OT constraintsystems (Stratal OT). Expressions are interpretedincrementally as they are built up, so morphology andphonology are intrinsically cyclic and local. (Interleavingnow also in DM, Embick 2010).
![Page 24: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/24.jpg)
The “derivational residue”
1 Two major problemsOpacityCyclicity
2 Two types of solutions
Introduce transderivational faithfulness constraints(Sympathy, O/O) and/or transderivational constraintsformulated over faithfulness relations (OT-CC).Modularity: level-ordered cascade of classic OT constraintsystems (Stratal OT). Expressions are interpretedincrementally as they are built up, so morphology andphonology are intrinsically cyclic and local. (Interleavingnow also in DM, Embick 2010).
![Page 25: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/25.jpg)
The “derivational residue”
1 Two major problemsOpacityCyclicity
2 Two types of solutions
Introduce transderivational faithfulness constraints(Sympathy, O/O) and/or transderivational constraintsformulated over faithfulness relations (OT-CC).Modularity: level-ordered cascade of classic OT constraintsystems (Stratal OT). Expressions are interpretedincrementally as they are built up, so morphology andphonology are intrinsically cyclic and local. (Interleavingnow also in DM, Embick 2010).
![Page 26: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/26.jpg)
The “derivational residue”
1 Two major problemsOpacityCyclicity
2 Two types of solutions
Introduce transderivational faithfulness constraints(Sympathy, O/O) and/or transderivational constraintsformulated over faithfulness relations (OT-CC).Modularity: level-ordered cascade of classic OT constraintsystems (Stratal OT). Expressions are interpretedincrementally as they are built up, so morphology andphonology are intrinsically cyclic and local. (Interleavingnow also in DM, Embick 2010).
![Page 27: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/27.jpg)
The “derivational residue”
1 Two major problemsOpacityCyclicity
2 Two types of solutions
Introduce transderivational faithfulness constraints(Sympathy, O/O) and/or transderivational constraintsformulated over faithfulness relations (OT-CC).Modularity: level-ordered cascade of classic OT constraintsystems (Stratal OT). Expressions are interpretedincrementally as they are built up, so morphology andphonology are intrinsically cyclic and local. (Interleavingnow also in DM, Embick 2010).
![Page 28: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/28.jpg)
The “derivational residue”
1 Two major problemsOpacityCyclicity
2 Two types of solutions
Introduce transderivational faithfulness constraints(Sympathy, O/O) and/or transderivational constraintsformulated over faithfulness relations (OT-CC).Modularity: level-ordered cascade of classic OT constraintsystems (Stratal OT). Expressions are interpretedincrementally as they are built up, so morphology andphonology are intrinsically cyclic and local. (Interleavingnow also in DM, Embick 2010).
![Page 29: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/29.jpg)
Andrew Nevins
• Locality in Vowel Harmony (2010)• (with Karlos Arregi) Morphotactics: Basque Auxiliaries and theStructure of Spellout (in press)
Leading ideas:
1 Locality, restrictiveness.
2 Phonetic grounding, naturalness, markedess.
3 Modularity: separate morphology and syntax.
4 Cross-modular structural parallelism: phonology is notdifferent.
![Page 30: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/30.jpg)
Andrew Nevins
• Locality in Vowel Harmony (2010)• (with Karlos Arregi) Morphotactics: Basque Auxiliaries and theStructure of Spellout (in press)
Leading ideas:
1 Locality, restrictiveness.
2 Phonetic grounding, naturalness, markedess.
3 Modularity: separate morphology and syntax.
4 Cross-modular structural parallelism: phonology is notdifferent.
![Page 31: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/31.jpg)
Andrew Nevins
• Locality in Vowel Harmony (2010)• (with Karlos Arregi) Morphotactics: Basque Auxiliaries and theStructure of Spellout (in press)
Leading ideas:
1 Locality, restrictiveness.
2 Phonetic grounding, naturalness, markedess.
3 Modularity: separate morphology and syntax.
4 Cross-modular structural parallelism: phonology is notdifferent.
![Page 32: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/32.jpg)
Andrew Nevins
• Locality in Vowel Harmony (2010)• (with Karlos Arregi) Morphotactics: Basque Auxiliaries and theStructure of Spellout (in press)
Leading ideas:
1 Locality, restrictiveness.
2 Phonetic grounding, naturalness, markedess.
3 Modularity: separate morphology and syntax.
4 Cross-modular structural parallelism: phonology is notdifferent.
![Page 33: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/33.jpg)
How to achieve this?
1 Nevins’ proposal
Rules plus (i) constraints that block rules, (ii) constraintsthat trigger rules.Functional submodules in the morphology: Feature andnode deletion → Linearization → Vocabulary Insertion →
Movement and Copying
2 Stratal OT
Only constraints, cyclically evaluated.Structural submodules in the morphology: Stems andWords.Beyond cross-modular parallelism: constraint-basedapproaches in phonetic implementation (Flemming),metrics, processing.
![Page 34: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/34.jpg)
How to achieve this?
1 Nevins’ proposal
Rules plus (i) constraints that block rules, (ii) constraintsthat trigger rules.Functional submodules in the morphology: Feature andnode deletion → Linearization → Vocabulary Insertion →
Movement and Copying
2 Stratal OT
Only constraints, cyclically evaluated.Structural submodules in the morphology: Stems andWords.Beyond cross-modular parallelism: constraint-basedapproaches in phonetic implementation (Flemming),metrics, processing.
![Page 35: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/35.jpg)
How to achieve this?
1 Nevins’ proposal
Rules plus (i) constraints that block rules, (ii) constraintsthat trigger rules.Functional submodules in the morphology: Feature andnode deletion → Linearization → Vocabulary Insertion →
Movement and Copying
2 Stratal OT
Only constraints, cyclically evaluated.Structural submodules in the morphology: Stems andWords.Beyond cross-modular parallelism: constraint-basedapproaches in phonetic implementation (Flemming),metrics, processing.
![Page 36: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/36.jpg)
How to achieve this?
1 Nevins’ proposal
Rules plus (i) constraints that block rules, (ii) constraintsthat trigger rules.Functional submodules in the morphology: Feature andnode deletion → Linearization → Vocabulary Insertion →
Movement and Copying
2 Stratal OT
Only constraints, cyclically evaluated.Structural submodules in the morphology: Stems andWords.Beyond cross-modular parallelism: constraint-basedapproaches in phonetic implementation (Flemming),metrics, processing.
![Page 37: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/37.jpg)
How to achieve this?
1 Nevins’ proposal
Rules plus (i) constraints that block rules, (ii) constraintsthat trigger rules.Functional submodules in the morphology: Feature andnode deletion → Linearization → Vocabulary Insertion →
Movement and Copying
2 Stratal OT
Only constraints, cyclically evaluated.Structural submodules in the morphology: Stems andWords.Beyond cross-modular parallelism: constraint-basedapproaches in phonetic implementation (Flemming),metrics, processing.
![Page 38: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/38.jpg)
How to achieve this?
1 Nevins’ proposal
Rules plus (i) constraints that block rules, (ii) constraintsthat trigger rules.Functional submodules in the morphology: Feature andnode deletion → Linearization → Vocabulary Insertion →
Movement and Copying
2 Stratal OT
Only constraints, cyclically evaluated.Structural submodules in the morphology: Stems andWords.Beyond cross-modular parallelism: constraint-basedapproaches in phonetic implementation (Flemming),metrics, processing.
![Page 39: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/39.jpg)
How to achieve this?
1 Nevins’ proposal
Rules plus (i) constraints that block rules, (ii) constraintsthat trigger rules.Functional submodules in the morphology: Feature andnode deletion → Linearization → Vocabulary Insertion →
Movement and Copying
2 Stratal OT
Only constraints, cyclically evaluated.Structural submodules in the morphology: Stems andWords.Beyond cross-modular parallelism: constraint-basedapproaches in phonetic implementation (Flemming),metrics, processing.
![Page 40: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/40.jpg)
Nevins’ procedural theory of vowel harmony
1 Search procedure: a value-seeking (“needy”) elementinitiates a search for the feature it needs, stops as soon asit finds the closest element bearing the relevant feature,and copies the value of that feature. If it can’t find a featurewithin the search domain, it defaults to a parametricallyspecified value.
2 Relativization parameter determines what values of theharmonic feature count as “relevant”: (a) all values, (b)contrastive values, (c) marked values.
3 Identity requirement may be imposed on the source andtarget of feature-copying.
4 Bounding parameters: (a) limits on search distance, (b)blocking by high-sonority elements.
![Page 41: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/41.jpg)
Nevins’ procedural theory of vowel harmony
1 Search procedure: a value-seeking (“needy”) elementinitiates a search for the feature it needs, stops as soon asit finds the closest element bearing the relevant feature,and copies the value of that feature. If it can’t find a featurewithin the search domain, it defaults to a parametricallyspecified value.
2 Relativization parameter determines what values of theharmonic feature count as “relevant”: (a) all values, (b)contrastive values, (c) marked values.
3 Identity requirement may be imposed on the source andtarget of feature-copying.
4 Bounding parameters: (a) limits on search distance, (b)blocking by high-sonority elements.
![Page 42: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/42.jpg)
Nevins’ procedural theory of vowel harmony
1 Search procedure: a value-seeking (“needy”) elementinitiates a search for the feature it needs, stops as soon asit finds the closest element bearing the relevant feature,and copies the value of that feature. If it can’t find a featurewithin the search domain, it defaults to a parametricallyspecified value.
2 Relativization parameter determines what values of theharmonic feature count as “relevant”: (a) all values, (b)contrastive values, (c) marked values.
3 Identity requirement may be imposed on the source andtarget of feature-copying.
4 Bounding parameters: (a) limits on search distance, (b)blocking by high-sonority elements.
![Page 43: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/43.jpg)
Nevins’ procedural theory of vowel harmony
1 Search procedure: a value-seeking (“needy”) elementinitiates a search for the feature it needs, stops as soon asit finds the closest element bearing the relevant feature,and copies the value of that feature. If it can’t find a featurewithin the search domain, it defaults to a parametricallyspecified value.
2 Relativization parameter determines what values of theharmonic feature count as “relevant”: (a) all values, (b)contrastive values, (c) marked values.
3 Identity requirement may be imposed on the source andtarget of feature-copying.
4 Bounding parameters: (a) limits on search distance, (b)blocking by high-sonority elements.
![Page 44: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/44.jpg)
An OT theory of vowel harmony
1 Phonology negotiates the conflicting claims of syntagmaticand paradigmatic markedness constraints and faithfulnessconstraints.
Harmony: *[αF][–αF]Paradigmatic markedness: *[μF]Faithfulness: IDENTSTEM(F), IDENT-σ1(F). . .
2 Constraints may be conjoined.
3 Cyclic evaluation.
Kiparsky & Pajusalu 2003
![Page 45: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/45.jpg)
An OT theory of vowel harmony
1 Phonology negotiates the conflicting claims of syntagmaticand paradigmatic markedness constraints and faithfulnessconstraints.
Harmony: *[αF][–αF]Paradigmatic markedness: *[μF]Faithfulness: IDENTSTEM(F), IDENT-σ1(F). . .
2 Constraints may be conjoined.
3 Cyclic evaluation.
Kiparsky & Pajusalu 2003
![Page 46: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/46.jpg)
An OT theory of vowel harmony
1 Phonology negotiates the conflicting claims of syntagmaticand paradigmatic markedness constraints and faithfulnessconstraints.
Harmony: *[αF][–αF]Paradigmatic markedness: *[μF]Faithfulness: IDENTSTEM(F), IDENT-σ1(F). . .
2 Constraints may be conjoined.
3 Cyclic evaluation.
Kiparsky & Pajusalu 2003
![Page 47: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/47.jpg)
An OT theory of vowel harmony
1 Phonology negotiates the conflicting claims of syntagmaticand paradigmatic markedness constraints and faithfulnessconstraints.
Harmony: *[αF][–αF]Paradigmatic markedness: *[μF]Faithfulness: IDENTSTEM(F), IDENT-σ1(F). . .
2 Constraints may be conjoined.
3 Cyclic evaluation.
Kiparsky & Pajusalu 2003
![Page 48: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/48.jpg)
An OT theory of vowel harmony
1 Phonology negotiates the conflicting claims of syntagmaticand paradigmatic markedness constraints and faithfulnessconstraints.
Harmony: *[αF][–αF]Paradigmatic markedness: *[μF]Faithfulness: IDENTSTEM(F), IDENT-σ1(F). . .
2 Constraints may be conjoined.
3 Cyclic evaluation.
Kiparsky & Pajusalu 2003
![Page 49: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/49.jpg)
An OT theory of vowel harmony
1 Phonology negotiates the conflicting claims of syntagmaticand paradigmatic markedness constraints and faithfulnessconstraints.
Harmony: *[αF][–αF]Paradigmatic markedness: *[μF]Faithfulness: IDENTSTEM(F), IDENT-σ1(F). . .
2 Constraints may be conjoined.
3 Cyclic evaluation.
Kiparsky & Pajusalu 2003
![Page 50: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/50.jpg)
Finnish
• järje-st-el-mä-llis-ty-ttä-mä-ttöm-yyde-llä-nsä-kään-kö-hän‘maybe not-even with his failure to have systematized?’
• suunn-it-el-ma-llis-tu-tta-ma-ttom-uude-lla-nsa-kaan-ko-han‘maybe not-even with his failure to have caused planning to beintroduced?’
• es-it-el-mä-llis-ty-ttä-mä-ttöm-yyde-llä-nsä-kään-kö-hän‘maybe not-even with his failure to have lecturing caused to beintroduced ?’
![Page 51: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/51.jpg)
The vowel system
u o a y ö ä i eBack + + + − − − − −
Round + + − + + − − −
High + − − + − − + −
Low − − + − − + − −
Harmony: u, o, a and y, ö, ä don’t co-occur.
Stems: pouta ‘fair weather’, pöytä ‘table’, *poutä, *poytä,*poyta, *pöuta. . .Suffixes: maa-ta ‘land’ (Part.Sg.), pää-tä ‘head’ (Part.Sg.)
![Page 52: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/52.jpg)
The vowel system
u o a y ö ä i eBack + + + − − − − −
Round + + − + + − − −
High + − − + − − + −
Low − − + − − + − −
Harmony: u, o, a and y, ö, ä don’t co-occur.
Stems: pouta ‘fair weather’, pöytä ‘table’, *poutä, *poytä,*poyta, *pöuta. . .Suffixes: maa-ta ‘land’ (Part.Sg.), pää-tä ‘head’ (Part.Sg.)
![Page 53: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/53.jpg)
The vowel system
u o a y ö ä i eBack + + + − − − − −
Round + + − + + − − −
High + − − + − − + −
Low − − + − − + − −
Harmony: u, o, a and y, ö, ä don’t co-occur.
Stems: pouta ‘fair weather’, pöytä ‘table’, *poutä, *poytä,*poyta, *pöuta. . .Suffixes: maa-ta ‘land’ (Part.Sg.), pää-tä ‘head’ (Part.Sg.)
![Page 54: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/54.jpg)
The neutral vowels i, e in Finnish
Unpaired in the underlying vowel inventory.Do not undergo suffixal harmony.
Transparent to suffixal harmony, e.g. tarina ‘tale’, tärinä‘vibration’ (*tarinä, *tärina)
Freely co-occur with back vowels in stems, e.g. piina‘torture’, viitta ‘cloak’
Trigger front harmony in suffixes, e.g. pii-nä ‘silicon’(Ess.Sg.), viit-tä ‘five’ (Part.Sg.)
![Page 55: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/55.jpg)
The neutral vowels i, e in Finnish
Unpaired in the underlying vowel inventory.Do not undergo suffixal harmony.
Transparent to suffixal harmony, e.g. tarina ‘tale’, tärinä‘vibration’ (*tarinä, *tärina)
Freely co-occur with back vowels in stems, e.g. piina‘torture’, viitta ‘cloak’
Trigger front harmony in suffixes, e.g. pii-nä ‘silicon’(Ess.Sg.), viit-tä ‘five’ (Part.Sg.)
![Page 56: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/56.jpg)
The neutral vowels i, e in Finnish
Unpaired in the underlying vowel inventory.Do not undergo suffixal harmony.
Transparent to suffixal harmony, e.g. tarina ‘tale’, tärinä‘vibration’ (*tarinä, *tärina)
Freely co-occur with back vowels in stems, e.g. piina‘torture’, viitta ‘cloak’
Trigger front harmony in suffixes, e.g. pii-nä ‘silicon’(Ess.Sg.), viit-tä ‘five’ (Part.Sg.)
![Page 57: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/57.jpg)
The neutral vowels i, e in Finnish
Unpaired in the underlying vowel inventory.Do not undergo suffixal harmony.
Transparent to suffixal harmony, e.g. tarina ‘tale’, tärinä‘vibration’ (*tarinä, *tärina)
Freely co-occur with back vowels in stems, e.g. piina‘torture’, viitta ‘cloak’
Trigger front harmony in suffixes, e.g. pii-nä ‘silicon’(Ess.Sg.), viit-tä ‘five’ (Part.Sg.)
![Page 58: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/58.jpg)
a, o, u, i, e are unmarked for [Back]
Text frequency of Finnish vowelsi 27 24.97%
a 23 22.88%e 16 15.49%u 10 11.91%o 10 10.67%ä 9 7.60%y 3 4.75%ö 1 1.78%
![Page 59: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/59.jpg)
Finnish constraint ranking
1 *1, *È
2 MARKEDHARMONY: *[αBack][–αBack] & *[μBack]: a domaincannot contain both a disharmonic vowel and a markedvowel.
3 FAITHFULNESS:
IDENTSTEM(BACK): An input [αBack] vowel in a Stem mustbe [αBack] in the output.IDENT-σ1(Back): An initial input [αBack] vowel must be[αBack] in the output.
4 HARMONY: *[αBack][–αBack]
This ranking gives Finnish, others generate an empiricallysupported factorial typology.
![Page 60: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/60.jpg)
Finnish constraint ranking
1 *1, *È
2 MARKEDHARMONY: *[αBack][–αBack] & *[μBack]: a domaincannot contain both a disharmonic vowel and a markedvowel.
3 FAITHFULNESS:
IDENTSTEM(BACK): An input [αBack] vowel in a Stem mustbe [αBack] in the output.IDENT-σ1(Back): An initial input [αBack] vowel must be[αBack] in the output.
4 HARMONY: *[αBack][–αBack]
This ranking gives Finnish, others generate an empiricallysupported factorial typology.
![Page 61: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/61.jpg)
Finnish constraint ranking
1 *1, *È
2 MARKEDHARMONY: *[αBack][–αBack] & *[μBack]: a domaincannot contain both a disharmonic vowel and a markedvowel.
3 FAITHFULNESS:
IDENTSTEM(BACK): An input [αBack] vowel in a Stem mustbe [αBack] in the output.IDENT-σ1(Back): An initial input [αBack] vowel must be[αBack] in the output.
4 HARMONY: *[αBack][–αBack]
This ranking gives Finnish, others generate an empiricallysupported factorial typology.
![Page 62: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/62.jpg)
Finnish constraint ranking
1 *1, *È
2 MARKEDHARMONY: *[αBack][–αBack] & *[μBack]: a domaincannot contain both a disharmonic vowel and a markedvowel.
3 FAITHFULNESS:
IDENTSTEM(BACK): An input [αBack] vowel in a Stem mustbe [αBack] in the output.IDENT-σ1(Back): An initial input [αBack] vowel must be[αBack] in the output.
4 HARMONY: *[αBack][–αBack]
This ranking gives Finnish, others generate an empiricallysupported factorial typology.
![Page 63: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/63.jpg)
Finnish constraint ranking
1 *1, *È
2 MARKEDHARMONY: *[αBack][–αBack] & *[μBack]: a domaincannot contain both a disharmonic vowel and a markedvowel.
3 FAITHFULNESS:
IDENTSTEM(BACK): An input [αBack] vowel in a Stem mustbe [αBack] in the output.IDENT-σ1(Back): An initial input [αBack] vowel must be[αBack] in the output.
4 HARMONY: *[αBack][–αBack]
This ranking gives Finnish, others generate an empiricallysupported factorial typology.
![Page 64: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/64.jpg)
Finnish constraint ranking
1 *1, *È
2 MARKEDHARMONY: *[αBack][–αBack] & *[μBack]: a domaincannot contain both a disharmonic vowel and a markedvowel.
3 FAITHFULNESS:
IDENTSTEM(BACK): An input [αBack] vowel in a Stem mustbe [αBack] in the output.IDENT-σ1(Back): An initial input [αBack] vowel must be[αBack] in the output.
4 HARMONY: *[αBack][–αBack]
This ranking gives Finnish, others generate an empiricallysupported factorial typology.
![Page 65: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/65.jpg)
Stems are subject only to M ARKED HARMONY
MARKEDHARMONY: a domain cannot contain both a markedvowel and a disharmonic vowel.
* [ aD äDM ]α✓ [ iD aD ]α✓ [ i äM ]α✓ [ aD iD aD ]α* [ aD iD äM ]α* [ äDM iD aD ]α✓ [ äM i äM ]α
![Page 66: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/66.jpg)
Suffixes undergo also H ARMONY
Input Candidates *1, *È MARKEDHARMONY IDENTSTEM(B) HARMONY
[ i ] a i a *☞ i ä
[ i a ] ☞ i a *i ä *
[ i ä ] i a * *☞ i ä
[ a i ] a ☞ a i a **a i ä * *a 1 a * *
a i a ☞ a i a **a i ä * *a 1 a * *
a i ä ☞ a i a **a i ä * *a 1 a * *
![Page 67: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/67.jpg)
Nevins
1 Harmony applies in suffixes and within roots.
2 A needy vowel seeks a contrastive feature to its left(‘needy’ ≈ ‘unspecified’).
3 Transparent i, e are excluded from the search.
4 If the search fails, default [–Back] is assigned.
5 Non-initial stem syllables also undergo harmony (evidencefrom language games). Disharmony handled by specifyingvowels as non-needy.
![Page 68: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/68.jpg)
Nevins
1 Harmony applies in suffixes and within roots.
2 A needy vowel seeks a contrastive feature to its left(‘needy’ ≈ ‘unspecified’).
3 Transparent i, e are excluded from the search.
4 If the search fails, default [–Back] is assigned.
5 Non-initial stem syllables also undergo harmony (evidencefrom language games). Disharmony handled by specifyingvowels as non-needy.
![Page 69: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/69.jpg)
Nevins
1 Harmony applies in suffixes and within roots.
2 A needy vowel seeks a contrastive feature to its left(‘needy’ ≈ ‘unspecified’).
3 Transparent i, e are excluded from the search.
4 If the search fails, default [–Back] is assigned.
5 Non-initial stem syllables also undergo harmony (evidencefrom language games). Disharmony handled by specifyingvowels as non-needy.
![Page 70: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/70.jpg)
Nevins
1 Harmony applies in suffixes and within roots.
2 A needy vowel seeks a contrastive feature to its left(‘needy’ ≈ ‘unspecified’).
3 Transparent i, e are excluded from the search.
4 If the search fails, default [–Back] is assigned.
5 Non-initial stem syllables also undergo harmony (evidencefrom language games). Disharmony handled by specifyingvowels as non-needy.
![Page 71: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/71.jpg)
Nevins
1 Harmony applies in suffixes and within roots.
2 A needy vowel seeks a contrastive feature to its left(‘needy’ ≈ ‘unspecified’).
3 Transparent i, e are excluded from the search.
4 If the search fails, default [–Back] is assigned.
5 Non-initial stem syllables also undergo harmony (evidencefrom language games). Disharmony handled by specifyingvowels as non-needy.
![Page 72: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/72.jpg)
Argument 1: stem harmony
1 Non-standard speakers nativize disharmonic stems:Peugeot → pösö, trotyyli → rotuli ‘TNT’, olympia- →olumpia-, pulityyri → pulituuri ‘furniture polish’.Never Kiina → *Kiinä ‘China’, metro → *metrö.
2 Ranking MH ≫ IDSTEM(B) ≫ H excludes Peugeot,trotyyli, but not Kiina, metro.
3 The search-and-copy approach doesn’t have a harmonyconstraint. It must specify the distribution of neediness by alookahead rule: “a stem vowel is needy only if its needs willbe satisfied by copying, not by default assignment.”
![Page 73: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/73.jpg)
Argument 1: stem harmony
1 Non-standard speakers nativize disharmonic stems:Peugeot → pösö, trotyyli → rotuli ‘TNT’, olympia- →olumpia-, pulityyri → pulituuri ‘furniture polish’.Never Kiina → *Kiinä ‘China’, metro → *metrö.
2 Ranking MH ≫ IDSTEM(B) ≫ H excludes Peugeot,trotyyli, but not Kiina, metro.
3 The search-and-copy approach doesn’t have a harmonyconstraint. It must specify the distribution of neediness by alookahead rule: “a stem vowel is needy only if its needs willbe satisfied by copying, not by default assignment.”
![Page 74: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/74.jpg)
Argument 1: stem harmony
1 Non-standard speakers nativize disharmonic stems:Peugeot → pösö, trotyyli → rotuli ‘TNT’, olympia- →olumpia-, pulityyri → pulituuri ‘furniture polish’.Never Kiina → *Kiinä ‘China’, metro → *metrö.
2 Ranking MH ≫ IDSTEM(B) ≫ H excludes Peugeot,trotyyli, but not Kiina, metro.
3 The search-and-copy approach doesn’t have a harmonyconstraint. It must specify the distribution of neediness by alookahead rule: “a stem vowel is needy only if its needs willbe satisfied by copying, not by default assignment.”
![Page 75: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/75.jpg)
Argument 2: i,e don’t trigger harmony frommonosyllabic C-roots
itk-u ‘crying’ itke-sk-el-y ‘crying’ (durative)hiill-os ‘embers’ piene-nn-ös ‘reduction’tek-o ‘deed’ tee+sk+ent+el+y ‘pretending’pit-uus ‘length’ pid-emm-yys ‘greater length’pien-uus ‘small size’ pien-emm-yys ‘smaller size’
Analysis: the minimal stem is disyllabic, so the Root+Suffixcombination is the first cyclic constituent.
• (itk-u)ω (like monomorphemic letku ‘hose’)• ((itk-esk-el)ω-y)ω
![Page 76: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/76.jpg)
Argument 3: feature-changing VH
Non-needy (fully specified) vowels can harmonize, as predictedby constraint-based theory.
moinen ‘such’ moinen has underlying [+Back] /o/kum+moinen ‘which kind of?’ [+Back] after [+Back]mim+moinen ‘what kind of’ [+Back] after neutral [–Back]täm+möinen ‘this kind of’ [–Back] after harmonic [–Back]
• Similar examples in Hungarian VH (Vago).• Consonant assimilation is also applicable to fully specifiedvowels (Wetzels & Mascaró 2001).• Constraint-based theory unifies VH with other assimilationprocesses.
![Page 77: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/77.jpg)
Argument 4: Seto/Võru
Harmony like Finnish, but with two extra vowels /1/, /È/
u o a 1 È ü ö ä i eBack + + + + + − − − − −
Round + + − − − + + − − −
High + − − + − + − − + −
Low − − + − − − − + − −
/1/ back harmonic, occurs only in initial syllables.
/È/ occurs in any syllable, reduced to [@] non-initially.
/e/ is a neutral in initial syllables, front harmonic elsewhere.
/o/ is opaque.
/ö/ occurs only in initial syllables.
![Page 78: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/78.jpg)
Argument 4: Seto/Võru
Harmony like Finnish, but with two extra vowels /1/, /È/
u o a 1 È ü ö ä i eBack + + + + + − − − − −
Round + + − − − + + − − −
High + − − + − + − − + −
Low − − + − − − − + − −
/1/ back harmonic, occurs only in initial syllables.
/È/ occurs in any syllable, reduced to [@] non-initially.
/e/ is a neutral in initial syllables, front harmonic elsewhere.
/o/ is opaque.
/ö/ occurs only in initial syllables.
![Page 79: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/79.jpg)
Argument 4: Seto/Võru
Harmony like Finnish, but with two extra vowels /1/, /È/
u o a 1 È ü ö ä i eBack + + + + + − − − − −
Round + + − − − + + − − −
High + − − + − + − − + −
Low − − + − − − − + − −
/1/ back harmonic, occurs only in initial syllables.
/È/ occurs in any syllable, reduced to [@] non-initially.
/e/ is a neutral in initial syllables, front harmonic elsewhere.
/o/ is opaque.
/ö/ occurs only in initial syllables.
![Page 80: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/80.jpg)
Argument 4: Seto/Võru
Harmony like Finnish, but with two extra vowels /1/, /È/
u o a 1 È ü ö ä i eBack + + + + + − − − − −
Round + + − − − + + − − −
High + − − + − + − − + −
Low − − + − − − − + − −
/1/ back harmonic, occurs only in initial syllables.
/È/ occurs in any syllable, reduced to [@] non-initially.
/e/ is a neutral in initial syllables, front harmonic elsewhere.
/o/ is opaque.
/ö/ occurs only in initial syllables.
![Page 81: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/81.jpg)
Argument 4: Seto/Võru
Harmony like Finnish, but with two extra vowels /1/, /È/
u o a 1 È ü ö ä i eBack + + + + + − − − − −
Round + + − − − + + − − −
High + − − + − + − − + −
Low − − + − − − − + − −
/1/ back harmonic, occurs only in initial syllables.
/È/ occurs in any syllable, reduced to [@] non-initially.
/e/ is a neutral in initial syllables, front harmonic elsewhere.
/o/ is opaque.
/ö/ occurs only in initial syllables.
![Page 82: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/82.jpg)
Argument 4: Seto/Võru
Harmony like Finnish, but with two extra vowels /1/, /È/
u o a 1 È ü ö ä i eBack + + + + + − − − − −
Round + + − − − + + − − −
High + − − + − + − − + −
Low − − + − − − − + − −
/1/ back harmonic, occurs only in initial syllables.
/È/ occurs in any syllable, reduced to [@] non-initially.
/e/ is a neutral in initial syllables, front harmonic elsewhere.
/o/ is opaque.
/ö/ occurs only in initial syllables.
![Page 83: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/83.jpg)
Initial /i/ and / 1/ are distinctive, / 1/ triggersharmony
1 s1na ‘word’, kl1bisÈ-ma ‘to rattle’sinä ‘you’, libise-mä ‘to flutter’, silmä ‘eye’s’, hinneq ‘fiber’ilma ‘without’, minnu ‘me’, hinnÈq ‘grade’*C1Cä, *C1Cö, *C1Ce. . .
2 MH predicts this. Search procedure has a problem withidentifying the source.
Search for distinctive values excludes grammatical CiCa,CiCu, CiCÈ.Search for marked values fails to exclude either *CuCä,*CaCü, *CÈCe. . . , or *CäCu, *CüCa, *CöCÈ. . . , bothungrammatical.
![Page 84: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/84.jpg)
Initial /i/ and / 1/ are distinctive, / 1/ triggersharmony
1 s1na ‘word’, kl1bisÈ-ma ‘to rattle’sinä ‘you’, libise-mä ‘to flutter’, silmä ‘eye’s’, hinneq ‘fiber’ilma ‘without’, minnu ‘me’, hinnÈq ‘grade’*C1Cä, *C1Cö, *C1Ce. . .
2 MH predicts this. Search procedure has a problem withidentifying the source.
Search for distinctive values excludes grammatical CiCa,CiCu, CiCÈ.Search for marked values fails to exclude either *CuCä,*CaCü, *CÈCe. . . , or *CäCu, *CüCa, *CöCÈ. . . , bothungrammatical.
![Page 85: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/85.jpg)
Initial /i/ and / 1/ are distinctive, / 1/ triggersharmony
1 s1na ‘word’, kl1bisÈ-ma ‘to rattle’sinä ‘you’, libise-mä ‘to flutter’, silmä ‘eye’s’, hinneq ‘fiber’ilma ‘without’, minnu ‘me’, hinnÈq ‘grade’*C1Cä, *C1Cö, *C1Ce. . .
2 MH predicts this. Search procedure has a problem withidentifying the source.
Search for distinctive values excludes grammatical CiCa,CiCu, CiCÈ.Search for marked values fails to exclude either *CuCä,*CaCü, *CÈCe. . . , or *CäCu, *CüCa, *CöCÈ. . . , bothungrammatical.
![Page 86: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/86.jpg)
Initial /i/ and / 1/ are distinctive, / 1/ triggersharmony
1 s1na ‘word’, kl1bisÈ-ma ‘to rattle’sinä ‘you’, libise-mä ‘to flutter’, silmä ‘eye’s’, hinneq ‘fiber’ilma ‘without’, minnu ‘me’, hinnÈq ‘grade’*C1Cä, *C1Cö, *C1Ce. . .
2 MH predicts this. Search procedure has a problem withidentifying the source.
Search for distinctive values excludes grammatical CiCa,CiCu, CiCÈ.Search for marked values fails to exclude either *CuCä,*CaCü, *CÈCe. . . , or *CäCu, *CüCa, *CöCÈ. . . , bothungrammatical.
![Page 87: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/87.jpg)
The domain of harmony is the prosodic word
Harmony (local cases) -t-Deletion in Part.Pl.[+Back] total hits -i-Del. total hits
kúvernemèntti 80.25% 3,595 100.00% 248káramèlli 71.44% 9,843 100.00% 112,100árkkitèhti 55.42% 39,978 100.00% 219,800hárakìri 20.11% 2,496 100.00% 1508kúriiri 100.00% 103,553 90.26% 7,747bákteeri 99.99% 65,498 37.69% 353,802fákiiri 99.89% 1,755 84.93% 3,532kálenteri 98.78% 1,541,814 52.99% 181,743ártikkeli 99.21% 2,380,926 16.78% 2,048,650
• V, V: lexical accents. Analysis: monomorphemic wordsconsisting of two full feet are optionally prosodic compounds,e.g. (kúverne)ω(mèntti)ω, (kára)ω(mèlli)ω (Kiparsky 2003).
![Page 88: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/88.jpg)
Conceptual advantages
1 The computation need not refer to “distinctiveness”, aninherently global property.
2 Unifies harmony with other assimilation processes.
3 Relies on independently motivated prosodic domains.
![Page 89: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/89.jpg)
Conceptual advantages
1 The computation need not refer to “distinctiveness”, aninherently global property.
2 Unifies harmony with other assimilation processes.
3 Relies on independently motivated prosodic domains.
![Page 90: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/90.jpg)
Conceptual advantages
1 The computation need not refer to “distinctiveness”, aninherently global property.
2 Unifies harmony with other assimilation processes.
3 Relies on independently motivated prosodic domains.
![Page 91: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/91.jpg)
Basque clitics
1 Basic order of clitics in the auxiliary is Abs – T – Dat – Erg,the reverse of the normal Subject – Indirect Object – DirectObject order of arguments.
2 T
T
T
Abs T Dat Erg
3 CASEALIGNMENT
A clitic C1 c-commands C2 iff C1’s Th-role outranks C2’sTh-role.
![Page 92: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/92.jpg)
Basque clitics
1 Basic order of clitics in the auxiliary is Abs – T – Dat – Erg,the reverse of the normal Subject – Indirect Object – DirectObject order of arguments.
2 T
T
T
Abs T Dat Erg
3 CASEALIGNMENT
A clitic C1 c-commands C2 iff C1’s Th-role outranks C2’sTh-role.
![Page 93: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/93.jpg)
Basque clitics
1 Basic order of clitics in the auxiliary is Abs – T – Dat – Erg,the reverse of the normal Subject – Indirect Object – DirectObject order of arguments.
2 T
T
T
Abs T Dat Erg
3 CASEALIGNMENT
A clitic C1 c-commands C2 iff C1’s Th-role outranks C2’sTh-role.
![Page 94: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/94.jpg)
Second position requirement
1 NONINITIALITY:T in a finite verb cannot be the leftmost morpheme withinthe word.
2 ENCLISIS:Clitics are adjoined to the right of their host T.
![Page 95: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/95.jpg)
Second position requirement
1 NONINITIALITY:T in a finite verb cannot be the leftmost morpheme withinthe word.
2 ENCLISIS:Clitics are adjoined to the right of their host T.
![Page 96: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/96.jpg)
NO
NIN
ITIA
LIT
Y
EN
CL
ISIS
CA
SE
AL
IGN
ME
NT
Pres + Abs.1Sg + Erg.2Sga. *a-t-su T-1Sg-2Sg *b. ☞ n-a-su 1Sg-T-2Sg *c. *n-su-a 1Sg-2Sg-T **d. *a-su-t T-2Sg-1Sg * *e. *s-a-t 2Sg-T-1Sg * *f. *s-n-a 2Sg-1Sg-T ** *
![Page 97: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/97.jpg)
Morphological dissimilation
Delete 1pl.Abs/1pl.Dat in context of 2.Erg (Ondarru)
1 *2/1PL
An auxiliary cannot contain both a first plural clitic and asecond person clitic.
2 MAX-2PA second person argument must correspond to a clitic (or:it must agree).
![Page 98: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/98.jpg)
Morphological dissimilation
Delete 1pl.Abs/1pl.Dat in context of 2.Erg (Ondarru)
1 *2/1PL
An auxiliary cannot contain both a first plural clitic and asecond person clitic.
2 MAX-2PA second person argument must correspond to a clitic (or:it must agree).
![Page 99: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/99.jpg)
Tense
1 The proclitic morpheme d(o)- is assumed to mark presentindicative (Trask 1977, 1997, Donohue 2004); the T head towhich it is attached is unspecified for tense.
2 A&N treat it as a featureless epenthetic clitic, inserted tosatisfy a morphological constraint which requires thatTense must not begin a word. d-insertion is bled by a rulewhich moves an ergative clitic to the beginning of theauxiliary (Enclitic Metathesis). By stipulation, EncliticMetathesis only applies in the past tense; so d-insertiononly applies in the present tense.
![Page 100: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/100.jpg)
Tense
1 The proclitic morpheme d(o)- is assumed to mark presentindicative (Trask 1977, 1997, Donohue 2004); the T head towhich it is attached is unspecified for tense.
2 A&N treat it as a featureless epenthetic clitic, inserted tosatisfy a morphological constraint which requires thatTense must not begin a word. d-insertion is bled by a rulewhich moves an ergative clitic to the beginning of theauxiliary (Enclitic Metathesis). By stipulation, EncliticMetathesis only applies in the past tense; so d-insertiononly applies in the present tense.
![Page 101: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/101.jpg)
do-su ‘Present-2.Sg.’ is multiply ambiguous
1 Su-kyou.Sg-E
gu- /0us-A
ikus-isee-Prf
do-suPresent-2Sg
(Present + Abs.1Pl + Erg.2Sg)
‘You(Sg) have seen us.’
2 Su-kyou.Sg-E
gu-rius-D
emo-ngive-Prf
do-suPresent-2Sg
(Present + Dat.1Pl + Erg.2Sg)
‘You(Sg) have given it to us.’
3 Gu-riWe-Dat
su- /0you.Sg.Abs
gusta-tenlike-Perf
do-suPres-2Sg
(Present + Abs.2Sg + Dat.1Pl)
‘We like you(Sg.)’
![Page 102: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/102.jpg)
do-su ‘Present-2.Sg.’ is multiply ambiguous
1 Su-kyou.Sg-E
gu- /0us-A
ikus-isee-Prf
do-suPresent-2Sg
(Present + Abs.1Pl + Erg.2Sg)
‘You(Sg) have seen us.’
2 Su-kyou.Sg-E
gu-rius-D
emo-ngive-Prf
do-suPresent-2Sg
(Present + Dat.1Pl + Erg.2Sg)
‘You(Sg) have given it to us.’
3 Gu-riWe-Dat
su- /0you.Sg.Abs
gusta-tenlike-Perf
do-suPres-2Sg
(Present + Abs.2Sg + Dat.1Pl)
‘We like you(Sg.)’
![Page 103: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/103.jpg)
do-su ‘Present-2.Sg.’ is multiply ambiguous
1 Su-kyou.Sg-E
gu- /0us-A
ikus-isee-Prf
do-suPresent-2Sg
(Present + Abs.1Pl + Erg.2Sg)
‘You(Sg) have seen us.’
2 Su-kyou.Sg-E
gu-rius-D
emo-ngive-Prf
do-suPresent-2Sg
(Present + Dat.1Pl + Erg.2Sg)
‘You(Sg) have given it to us.’
3 Gu-riWe-Dat
su- /0you.Sg.Abs
gusta-tenlike-Perf
do-suPres-2Sg
(Present + Abs.2Sg + Dat.1Pl)
‘We like you(Sg.)’
![Page 104: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/104.jpg)
dosu ‘������X
XXXXXwe-Dat (like) you-Abs’ (Present)
NO
NIN
ITIA
LIT
Y
EN
CL
ISIS
CA
SE
AL
IGN
ME
NT
*2/1
PL
MA
X-2
P
FA
ITH
TE
NS
E
Present + Abs.2Sg + Dat.1Pl → d-o-su1a. *s-a-sku 2Sg-T-Dat.1Pl * * *1b. *a-sku T-Dat.1Pl * * *1c. *a-su T-2Sg * *1d. ☞ do-su Pres-T-2Sg *1e. *sku-a Dat.1Pl-T * * *1f. *s-a 2Sg-T * *1g. *do-su-a Pres-2Sg-T **1h. *do-sku-su Pres-T-Dat.1Pl-2Sg * * *
![Page 105: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/105.jpg)
sendun ‘������X
XXXXXwe-Dat (like)-d you-Abs’
NO
NIN
ITIA
LIT
Y
EN
CL
ISIS
*2/1
PL
MA
X-2
P
FA
ITH
TE
NS
E
Past + Abs.2Sg + Dat.1Pl → s-endu-n1a. *s-endu-sku-n 2Sg-T-Dat.1Pl * *1b. *endu-sku-n Past-Dat.1Pl * *1c. *endu-su-n Past-2Sg *1d. *d-endu-su-n Pres-Past-2Sg * *1e. ☞ s-endu-n 2Sg-Past *1f. *d-endu-sku-n Pres-Past-Dat.1Pl * * *
![Page 106: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/106.jpg)
sendu(n) ‘you-Erg (saw)������X
XXXXX
us-Abs’
NO
NIN
ITIA
LIT
Y
EN
CL
ISIS
*2/1
PL
MA
X-2
P
FA
ITH
TE
NS
E
Past + Abs.1Pl + Erg.2Sg → s-endu(-n) ‘you -ed us’1a. *g-endu-su-n 1Pl-Past-2Sg * *1b. *endu-su-n Past-2Sg *1c. *endu-gu-n Past-1Pl * *1d. *d-endu-su-n Pres-Past-2Sg * *1e. ☞ s-endu-n 2Sg-Past *1f. *g-endu-n 1Pl-Past * *1g. *d-endu-gu-n Pres-Past-1Pl * * *
![Page 107: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/107.jpg)
![Page 108: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/108.jpg)
y is translucent
Elative -sta/-stä, Inessive -ssa/-ssä, Allative -lla/-llä, Ablative-lta/-ltä. Google hits from Finnish pages.
[+Back] total hitstrotyyli 58.71% 1,669marttyyri 56.35% 3,110vampyyri 44.06% 32,692kalkyyli 20.66% 1,113analyysi 17.65% 1,414,089karikatyyri 4.63 % 9,572
![Page 109: Derivations - Massachusetts Institute of Technologyling50.mit.edu/wp-content/uploads/Kiparsky-slides.pdf · Derivations in phonology: a brief history 1 Kisseberth’s translation](https://reader033.vdocuments.net/reader033/viewer/2022060318/5f0c91107e708231d4360b4f/html5/thumbnails/109.jpg)
Other vowels are opaque
[+Back] total hitsmonttööri 0.00% 142jonglööri 0.04% 970amatööri 0.27% 68,941kuvernööri 0.01% 10,234miljardööri 0.00% 14,553vulgääri 1.02% 683afääri 0.79% 511karriääri 0.24% 837atmosfääri 0.05% 18,819miljonääri 0.00% 33,532syaani 100.00% 2,027tyranni 99.98% 11,730