Dresher (2009:§7.2.1) discusses an interesting morphophonological puzzle from the Inuit (Canadian) and Inupiaq (Alaskan) dialects of Eskimo-Aleut. These dialects descend from a four-phoneme vowel system *i, *u, *ə, *a, but in most dialects *ə has merged into *i, yielding three surface vowels: [i, a, u]. However, in some dialects (including Barrow Inupiaq), there appears to be a covert contrast between two “flavors” of i: “strong i” triggers palatalization of a following coronal consonant whereas “weak i” does not.

(1) Barrow Inupiaq (Kaplan 1981:§3.22, his 27-29):

a. iglu ‘house’, iglulu ‘and a house’, iglunik ‘houses’
b. ini ‘place’, inilu ‘and a place’, ‘ininik’ places’
c. iki ‘wound’, ikiʎu ‘and a wound’, ikiɲik ‘wounds’

Presumably, the stem-final i in (1b) is weak and the one in (1c) is strong.

Following some prior work, Dresher supposes that there is an underlying contrast between weak and strong i. He posits the following featural specification:¹

(2) Features for Barrow Inupiaq vowels (to be revised):

strong i: [Coronal, -Low]
weak i: [-Low]
/u/: [Labial, -Low]
/a/: [+Low]

This analysis has a close relationship to the theory of underspecification used in Logical Substance-Free Phonology (henceforth, LP);  I assume familiarity with the assumptions and operations of that theory, which have been discussed at length by Reiss and colleagues (Bale et al. 2020, Reiss 2021), including in an introductory textbook (Bale & Reiss 2018). Just a few modifications are needed, however.

For Dresher, who hypothesizes that non-contrastive features (computed using an algorithm he describes in detail; op. cit.:16) are the only ones which are phonologically active, it is not clear why strong i palatalizes coronal segments: clearly it is not that they are spreading the privative [Coronal], since that certainly would not trigger palatalization! One could, of course, adopt an analysis in which palatalization is not assimilatory. Alternatively, we could identify another feature specification which is characteristic of i and which might trigger palatalization of coronal consonants. Let us suppose this is in fact just [Palatal].² This gives us the following minimally-modified feature specification:

(2) Features for Barrow Inupiaq vowels (revised):

strong i: [Palatal, -Low]
weak i: [-Low]
/u/: [Labial, -Low]
/a/: [+Low]

According to Kaplan (§1.2), there are both plain and palatal coronal phonemes, so this seems to be a feature-changing process. Following the assumptions of LP that feature-changing processes derive from a deletion rule followed by an insertion rule, two rules are needed here; we give these below.

(3) [+Consonantal] \ {Coronal} / [Palatal] __
(4) [+Consonantal] ⊔ {Palatal} / [Palatal] __

Crucially, vowels other than strong i lack the [Palatal] specification to trigger (3-4).

Endnotes

1. Dresher’s analysis assumes privative features, but he notes elsewhere in the book that he usually adopts the features of his sources unless there is some relevant reason to dispute them.
2. If preferred, it is easy to translate the proposed analysis into one in which palatals are [-Back] and plain coronals are [+Back], à la Padgett (2003).

References

Bale, A., Papillon, M., and Reiss, C. 2014. Targeting underspecified segments: a formal analysis of feature-changing and feature-filling rules. Lingua 148: 240-253.
Bale, A., and Reiss, C. 2018. Phonology: a Formal Introduction. MIT Press.
Dresher, B. E. 2009. The Contrastive Hierarchy in Phonology. Cambridge University Press.
Kaplan, L. D. 1981. Phonological issues in North Alaskan Inupiaq. Alaska Native Language Center.
Padgett, J. 2003. Contrast and post-velar fronting in Russian. Natural Language and Linguistic Theory 21: 39-87.
Reiss, C. 2021. Towards a complete Logical Phonology model of intrasegmental changes. Glossa 6:107.