Two Allomorphy Models without Late Insertion (Morphology as Syntax)

 Here is a handout from my seminar with Richard Kayne (Morphology as Syntax). On this handout, I outline two approaches to allomorphy not involving late insertion. This list is not meant to be exclusive. In fact, in the very same class period (September 23 2023), I presented yet another completely different model of allomorphy based on autosegmental phonology.

Allomorphy Model 1 (Inner and Outer Morphemes)

[From Collins and Kayne 2023]

1.

There are two affixes AF1 and AF2. (e.g., -en plural and -s plural)

2.

AF1 is irregular and AF2 is regular.

AF1 selects particular roots. AF2 selects general category (e.g., NP).

3.

General Principle (UG): Irregular affixes are closer to the root than regular affixes.

4.

AF2 has formal feature F2 (e.g., PL). (e.g., -s plural)

AF1 has formal feature F2         (e.g., -en plural)

F1 and F2 both are semantically plural.

5.

Regular Affixation: X-AF2 (e.g., dog-s)

6.

Irregular Affixation: Y-AF1 (e.g., ox-en)

7.

Full Interpretation: *Y-AF1-AF2

Allomorphy Model 2 (Stacking and Deletion)

[From Collins 2018]

1.

There are two affixes AF1 and AF2. (e.g., -en plural and -s plural)

2.

AF1 is irregular and AF2 is regular.

AF1 selects particular roots. AF2 selects general category (e.g., NP).

3.

General Principle (UG): Irregular affixes are closer to the root than regular affixes.

4.

AF2 has formal feature F (e.g., PL). (e.g., -s plural)

AF1 has formal feature [uF] (e.g., [uPL]) (e.g., -en plural)

5.

Regular Affixation: X-AF2 (e.g., dog-s)

6.

Irregular Affixation: *Y-AF1 (AF1 is uninterpretable)

7.

Irregular Affixation: Y-AF1-AF2

8.

AF1 deletes AF2 by ICP: Y-AF1-AF2 --> Y-AF1-∅

(e.g., ox-en-s --> ox-en-∅)

9.

Invisible Category Principle (Emonds 1987) (UG)

A closed category B with positively specified features C may remain empty throughout a syntactic derivation if the features C (save possibly B itself) all alternatively realized in a phrasal sister of B.