LABIAL PLACE IN PHONOLOGY: UNIVERSAL AND VARIABLE Daniel Currie - PowerPoint PPT Presentation

Inventories show a tendency to avoid marked feature values. is looks circular. (Languages tend to avoid rare sounds?) But it does have empirical content… TYPOLOGICAL EXPECTATIONS Marked feature avoidance Clements (2009: 42): Marked feature values can be defined as those that are not present in all languages. 10 / 42

is looks circular. (Languages tend to avoid rare sounds?) But it does have empirical content… TYPOLOGICAL EXPECTATIONS Marked feature avoidance Clements (2009: 42): Marked feature values can be defined as those that are not present in all languages. Inventories show a tendency to avoid marked feature values. 10 / 42

But it does have empirical content… TYPOLOGICAL EXPECTATIONS Marked feature avoidance Clements (2009: 42): Marked feature values can be defined as those that are not present in all languages. Inventories show a tendency to avoid marked feature values. is looks circular. (Languages tend to avoid rare sounds?) 10 / 42

TYPOLOGICAL EXPECTATIONS Marked feature avoidance Clements (2009: 42): Marked feature values can be defined as those that are not present in all languages. Inventories show a tendency to avoid marked feature values. is looks circular. (Languages tend to avoid rare sounds?) But it does have empirical content… 10 / 42

Marked sounds are a last (or at least laer) resort for any inventory. ere should be no sounds that occur only in small inventories. e markedness of a segment should correlate negatively with the number of inventories in which it occurs… …and positively with their mean size. TYPOLOGICAL EXPECTATIONS Marked feature avoidance Predictions of Marked feature avoidance: Paerns of markedness should hold both within and between languages. 11 / 42

ere should be no sounds that occur only in small inventories. e markedness of a segment should correlate negatively with the number of inventories in which it occurs… …and positively with their mean size. TYPOLOGICAL EXPECTATIONS Marked feature avoidance Predictions of Marked feature avoidance: Paerns of markedness should hold both within and between languages. Marked sounds are a last (or at least laer) resort for any inventory. 11 / 42

e markedness of a segment should correlate negatively with the number of inventories in which it occurs… …and positively with their mean size. TYPOLOGICAL EXPECTATIONS Marked feature avoidance Predictions of Marked feature avoidance: Paerns of markedness should hold both within and between languages. Marked sounds are a last (or at least laer) resort for any inventory. ere should be no sounds that occur only in small inventories. 11 / 42

…and positively with their mean size. TYPOLOGICAL EXPECTATIONS Marked feature avoidance Predictions of Marked feature avoidance: Paerns of markedness should hold both within and between languages. Marked sounds are a last (or at least laer) resort for any inventory. ere should be no sounds that occur only in small inventories. e markedness of a segment should correlate negatively with the number of inventories in which it occurs… 11 / 42

TYPOLOGICAL EXPECTATIONS Marked feature avoidance Predictions of Marked feature avoidance: Paerns of markedness should hold both within and between languages. Marked sounds are a last (or at least laer) resort for any inventory. ere should be no sounds that occur only in small inventories. e markedness of a segment should correlate negatively with the number of inventories in which it occurs… …and positively with their mean size. 11 / 42

450 Number of inventories in UPSID containing segments of the indicated type 69 68 23 . . . . . . . . K Kʷ K’ Kʷ’ K = plain dorsal stop; Kʷ= labialized dorsal stop; K’ = plain dorsal ejective; Kʷ’ = labialized dorsal ejective TYPOLOGICAL EXPECTATIONS Marked feature avoidance Clements (2009: 42): Marked segments occur in fewer inventories. 12 / 42

450 69 68 23 . . . . . . . . K Kʷ K’ Kʷ’ K = plain dorsal stop; Kʷ= labialized dorsal stop; K’ = plain dorsal ejective; Kʷ’ = labialized dorsal ejective TYPOLOGICAL EXPECTATIONS Marked feature avoidance Clements (2009: 42): Marked segments occur in fewer inventories. Number of inventories in UPSID containing segments of the indicated type 12 / 42

TYPOLOGICAL EXPECTATIONS Marked feature avoidance Clements (2009: 42): Marked segments occur in fewer inventories. 450 Number of inventories in UPSID containing segments of the indicated type 69 68 23 . . . . . . . . . K Kʷ K’ Kʷ’ K = plain dorsal stop; Kʷ= labialized dorsal stop; K’ = plain dorsal ejective; Kʷ’ = labialized dorsal ejective 12 / 42

Mean number of consonants in UPSID inventories 35.8 containing segments of the indicated type 29.0 26.4 19.7 . . . . . . . . K Kʷ K’ Kʷ’ K = plain dorsal stop; Kʷ= labialized dorsal stop; K’ = plain dorsal ejective; Kʷ’ = labialized dorsal ejective TYPOLOGICAL EXPECTATIONS Marked feature avoidance Clements (2009: 42): Marked segments occur in larger inventories. 13 / 42

35.8 29.0 26.4 19.7 . . . . . . . . K Kʷ K’ Kʷ’ K = plain dorsal stop; Kʷ= labialized dorsal stop; K’ = plain dorsal ejective; Kʷ’ = labialized dorsal ejective TYPOLOGICAL EXPECTATIONS Marked feature avoidance Clements (2009: 42): Marked segments occur in larger inventories. Mean number of consonants in UPSID inventories containing segments of the indicated type 13 / 42

TYPOLOGICAL EXPECTATIONS Marked feature avoidance Clements (2009: 42): Marked segments occur in larger inventories. Mean number of consonants in UPSID inventories 35.8 containing segments of the indicated type 29.0 26.4 19.7 . . . . . . . . . K Kʷ K’ Kʷ’ K = plain dorsal stop; Kʷ= labialized dorsal stop; K’ = plain dorsal ejective; Kʷ’ = labialized dorsal ejective 13 / 42

In Unified Feature eory (Clements & Hume 1995): k kʷ k͡p C C C C-place C-place C-place [dorsal] [dorsal] vocalic [dorsal] [labial] V-place [labial] TYPOLOGICAL EXPECTATIONS The representation of labialization Clements (2009): A marked feature [+rounded] distinguishes Kʷ. 14 / 42

k kʷ k͡p C C C C-place C-place C-place [dorsal] [dorsal] vocalic [dorsal] [labial] V-place [labial] TYPOLOGICAL EXPECTATIONS The representation of labialization Clements (2009): A marked feature [+rounded] distinguishes Kʷ. In Unified Feature eory (Clements & Hume 1995): 14 / 42

kʷ k͡p C C C-place C-place [dorsal] vocalic [dorsal] [labial] V-place [labial] TYPOLOGICAL EXPECTATIONS The representation of labialization Clements (2009): A marked feature [+rounded] distinguishes Kʷ. In Unified Feature eory (Clements & Hume 1995): k C C-place [dorsal] 14 / 42

k͡p C C-place [dorsal] [labial] The representation of labialization TYPOLOGICAL EXPECTATIONS Clements (2009): A marked feature [+rounded] distinguishes Kʷ. In Unified Feature eory (Clements & Hume 1995): k kʷ C C C-place C-place ❚ ❥❥❥❥❥❥❥ ❚ ❚ ❚ ❚ ❚ ❚ [dorsal] [dorsal] vocalic V-place [labial] 14 / 42

TYPOLOGICAL EXPECTATIONS The representation of labialization Clements (2009): A marked feature [+rounded] distinguishes Kʷ. In Unified Feature eory (Clements & Hume 1995): k kʷ k͡p C C C C-place C-place C-place ❚ ❚ ❥❥❥❥❥❥❥ ❚ ❥❥❥❥❥❥❥ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ [dorsal] [dorsal] vocalic [dorsal] [labial] V-place [labial] 14 / 42

e presence of labialization on consonants is marked… …but there is no reason to expect any specific combination of primary and secondary place to be more or less marked than the others… …except to the extent that some place features are inherently more or less marked than others. TYPOLOGICAL EXPECTATIONS The representation of labialization Expectations: Primary and secondary articulations can vary independently (within anatomical limits). 15 / 42

…but there is no reason to expect any specific combination of primary and secondary place to be more or less marked than the others… …except to the extent that some place features are inherently more or less marked than others. TYPOLOGICAL EXPECTATIONS The representation of labialization Expectations: Primary and secondary articulations can vary independently (within anatomical limits). e presence of labialization on consonants is marked… 15 / 42

…except to the extent that some place features are inherently more or less marked than others. TYPOLOGICAL EXPECTATIONS The representation of labialization Expectations: Primary and secondary articulations can vary independently (within anatomical limits). e presence of labialization on consonants is marked… …but there is no reason to expect any specific combination of primary and secondary place to be more or less marked than the others… 15 / 42

TYPOLOGICAL EXPECTATIONS The representation of labialization Expectations: Primary and secondary articulations can vary independently (within anatomical limits). e presence of labialization on consonants is marked… …but there is no reason to expect any specific combination of primary and secondary place to be more or less marked than the others… …except to the extent that some place features are inherently more or less marked than others. 15 / 42

. TYPOLOGICAL SURPRISES . TYPOLOGICAL SURPRISES 16 / 42

TYPOLOGICAL SURPRISES Database: P-base (Mielke 2008) 628 varieties of 548 spoken languages 17 / 42

primary place manner (except nasality) airstream Labialization largely cross-classifies with: Unsurprising inventories TYPOLOGICAL SURPRISES Tangale (Chadic) p t k ʔ tʷ kʷ b d̪ d ʤ ɡ bʷ dʷ ɡʷ ᵐb ⁿd ⁿʤ ฀ɡ ɓ ɗ ɓʷ ɗʷ s ʃ sʷ ʃʷ z ʒ zʷ ʒʷ m n ŋ w r j l rʷ jʷ 18 / 42

primary place manner (except nasality) airstream TYPOLOGICAL SURPRISES Tangale (Chadic) Unsurprising inventories Labialization largely cross-classifies with: p t k ʔ tʷ kʷ b d̪ d ʤ ɡ bʷ dʷ ɡʷ ᵐb ⁿd ⁿʤ ฀ɡ ɓ ɗ ɓʷ ɗʷ s ʃ sʷ ʃʷ z ʒ zʷ ʒʷ m n ŋ w r j l rʷ jʷ 18 / 42

manner (except nasality) airstream TYPOLOGICAL SURPRISES Tangale (Chadic) Unsurprising inventories Labialization largely cross-classifies with: p t k ʔ primary place tʷ kʷ b d̪ d ʤ ɡ bʷ dʷ ɡʷ ᵐb ⁿd ⁿʤ ฀ɡ ɓ ɗ ɓʷ ɗʷ s ʃ sʷ ʃʷ z ʒ zʷ ʒʷ m n ŋ w r j l rʷ jʷ 18 / 42

airstream TYPOLOGICAL SURPRISES Tangale (Chadic) Unsurprising inventories Labialization largely cross-classifies with: p t k ʔ primary place tʷ kʷ manner (except nasality) b d̪ d ʤ ɡ bʷ dʷ ɡʷ ᵐb ⁿd ⁿʤ ฀ɡ ɓ ɗ ɓʷ ɗʷ s ʃ sʷ ʃʷ z ʒ zʷ ʒʷ m n ŋ w r j l rʷ jʷ 18 / 42

TYPOLOGICAL SURPRISES Tangale (Chadic) Unsurprising inventories Labialization largely cross-classifies with: p t k ʔ primary place tʷ kʷ manner (except nasality) b d̪ d ʤ ɡ bʷ dʷ ɡʷ airstream ᵐb ⁿd ⁿʤ ฀ɡ ɓ ɗ ɓʷ ɗʷ s ʃ sʷ ʃʷ z ʒ zʷ ʒʷ m n ŋ w r j l rʷ jʷ 18 / 42

Secondary labialization is contrastive only for /k/–/kʷ/. ere are no consonantal segments with primary labial place. Economy: Minimal benefit from [±round]. Markedness: More marked /kʷ/ should entail less marked /p/. TYPOLOGICAL SURPRISES Surprising inventories Wichita (Caddoan) t k ʔ kʷ ʦ s h r j w 19 / 42

ere are no consonantal segments with primary labial place. Economy: Minimal benefit from [±round]. Markedness: More marked /kʷ/ should entail less marked /p/. Surprising inventories TYPOLOGICAL SURPRISES Wichita (Caddoan) Secondary labialization is contrastive only for t k ʔ /k/–/kʷ/. kʷ ʦ s h r j w 19 / 42

Economy: Minimal benefit from [±round]. Markedness: More marked /kʷ/ should entail less marked /p/. Surprising inventories TYPOLOGICAL SURPRISES Wichita (Caddoan) Secondary labialization is contrastive only for t k ʔ /k/–/kʷ/. kʷ ere are no consonantal ʦ segments with primary s h labial place. r j w 19 / 42

Markedness: More marked /kʷ/ should entail less marked /p/. Surprising inventories TYPOLOGICAL SURPRISES Wichita (Caddoan) Secondary labialization is contrastive only for t k ʔ /k/–/kʷ/. kʷ ere are no consonantal ʦ segments with primary s h labial place. r Economy: Minimal benefit j w from [±round]. 19 / 42

TYPOLOGICAL SURPRISES Wichita (Caddoan) Surprising inventories Secondary labialization is contrastive only for t k ʔ /k/–/kʷ/. kʷ ere are no consonantal ʦ segments with primary s h labial place. r Economy: Minimal benefit j w from [±round]. Markedness: More marked /kʷ/ should entail less marked /p/. 19 / 42

26 have only one labialized consonant. TYPOLOGICAL SURPRISES Surprising inventories 117 inventories in P-base have at least one labialized consonant. 20 / 42

TYPOLOGICAL SURPRISES Surprising inventories 117 inventories in P-base have at least one labialized consonant. 26 have only one labialized consonant. 20 / 42

TYPOLOGICAL SURPRISES Surprising inventories Cuna (Chibchan) 117 inventories in P-base have at least one labialized consonant. 26 have only one labialized consonant. p t k kʷ s m n l r w j 20 / 42

TYPOLOGICAL SURPRISES Surprising inventories Cuna (Chibchan) Passamaquoddy (Algic) 117 inventories in P-base have at least one labialized consonant. 26 have only one labialized consonant. p t k p t ʧ k kʷ kʷ s s h m n m n l l r w j w j 20 / 42

TYPOLOGICAL SURPRISES Surprising inventories Comanche (Uto-Aztecan) Passamaquoddy (Algic) 117 inventories in P-base have at least one labialized consonant. 26 have only one labialized consonant. p t ʧ k p t k ʔ kʷ s h kʷ ʦ m n l s h m n w j w j 20 / 42

TYPOLOGICAL SURPRISES Surprising inventories Comanche (Uto-Aztecan) Dani (Austronesian) 117 inventories in P-base have at least one labialized consonant. 26 have only one labialized consonant. p t k ʔ p t k ʔ kʷ s h kʷ ʦ m n ŋ l s h m n w j w j 20 / 42

TYPOLOGICAL SURPRISES Esse Ejja (Tacanan) Dani (Austronesian) Surprising inventories 117 inventories in P-base have at least one labialized consonant. 26 have only one labialized consonant. p t ʧ k ʔ p t k ʔ kʷ kʷ ɓ ɗ s h s ʃ x h m n ŋ m n ɲ l w j w j 20 / 42

*kʷ (Dedrick & Casad 1999) TYPOLOGICAL SURPRISES Esse Ejja (Tacanan) Sonora Yaqui (Uto-Aztecan) Surprising inventories 117 inventories in P-base have at least one labialized consonant. 26 have only one labialized consonant. p t ʧ k ʔ p t ʧ k ʔ kʷ b ɓ ɗ bʷ s ʃ x h s h m n ɲ m n w j l ɾ w j 20 / 42

TYPOLOGICAL SURPRISES Surprising inventories Esse Ejja (Tacanan) Sonora Yaqui (Uto-Aztecan) 117 inventories in P-base have at least one labialized consonant. 26 have only one labialized consonant. p t ʧ k ʔ p t ʧ k ʔ kʷ b ɓ ɗ bʷ < *kʷ (Dedrick & Casad 1999) s ʃ x h s h m n ɲ m n w j l ɾ w j 20 / 42

TYPOLOGICAL SURPRISES Surprising inventories Even in inventories with more labialized consonants, rounding oen does not cross-classify with place: 21 / 42

TYPOLOGICAL SURPRISES Surprising inventories Kombai (Trans-New Guinea) Even in inventories with more labialized consonants, rounding oen does not cross-classify with place: ᵐb ⁿd ɟ ฀ɡ ฀ɡʷ ɸ x xʷ l r j ɥ w 21 / 42

TYPOLOGICAL SURPRISES Ojibwa (Algic) Kombai (Trans-New Guinea) Surprising inventories Even in inventories with more labialized consonants, rounding oen does not cross-classify with place: ᵐb ⁿd ɟ ฀ɡ ฀ɡʷ p t ʧ k kʷ ɸ x xʷ b d ʤ ɡ ɡʷ l s ʃ h hʷ r z ʒ j ɥ w m n ŋ j w 21 / 42

TYPOLOGICAL SURPRISES Ojibwa (Algic) Sinaugoro (Austronesian) Surprising inventories Even in inventories with more labialized consonants, rounding oen does not cross-classify with place: t k kʷ p t ʧ k kʷ b d ɡ ɡʷ b d ʤ ɡ ɡʷ f s s ʃ h hʷ v r ɣ ɣʷ z ʒ m n m n ŋ l j w 21 / 42

TYPOLOGICAL SURPRISES Tigrinya (Semitic) Sinaugoro (Austronesian) Surprising inventories Even in inventories with more labialized consonants, rounding oen does not cross-classify with place: t k kʷ p t ʧ k kʷ ʔ b d ɡ ɡʷ b d ʤ ɡ ɡʷ f s p’ t’ ʧ’ k’ kʷ’ v r ɣ ɣʷ f s ʃ ħ h m n z ʒ ʕ l s’ m n ɲ r l j w 21 / 42

TYPOLOGICAL SURPRISES Tigrinya (Semitic) Halkomelem (Salishan) Surprising inventories Even in inventories with more labialized consonants, rounding oen does not cross-classify with place: p t k kʷ q qʷ ʔ p t ʧ k kʷ ʔ p’ t’ k’ k’ʷ q’ q’ʷ b d ʤ ɡ ɡʷ ʧ p’ t’ ʧ’ k’ kʷ’ tθ’ tɬ’ ʧ’ f s ʃ ħ h θ ɬ ʃ ç x xʷ χ χʷ h z ʒ ʕ m s’ l m n ɲ j w r l j w 21 / 42

TYPOLOGICAL SURPRISES Kilivila (Austronesian) Halkomelem (Salishan) Surprising inventories Even in inventories with more labialized consonants, rounding oen does not cross-classify with place: p t k kʷ q qʷ ʔ p pʷ t k kʷ p’ t’ k’ k’ʷ q’ q’ʷ b bʷ d ɡ ɡʷ ʧ β s tθ’ tɬ’ ʧ’ m mʷ n ŋ θ ɬ ʃ ç x xʷ χ χʷ h r m l l j w j w 21 / 42

585 100 6 . . . . . . t kʷ tʷ Number of inventories in 584 P-base containing the indicated segments . . k TYPOLOGICAL SURPRISES Surprising numbers Applying Clements’s (2009) diagnostics for markedness: 22 / 42

585 100 6 . . . . . . t kʷ tʷ 584 . . k TYPOLOGICAL SURPRISES Surprising numbers Applying Clements’s (2009) diagnostics for markedness: Number of inventories in P-base containing the indicated segments 22 / 42

585 100 6 . . . . . . t kʷ tʷ Surprising numbers TYPOLOGICAL SURPRISES Applying Clements’s (2009) diagnostics for markedness: Number of inventories in 584 P-base containing the indicated segments . . . k 22 / 42

100 6 . . . . kʷ tʷ TYPOLOGICAL SURPRISES Surprising numbers Applying Clements’s (2009) diagnostics for markedness: Number of inventories in 584 585 P-base containing the indicated segments . . . . . t k 22 / 42

6 . . tʷ TYPOLOGICAL SURPRISES Surprising numbers Applying Clements’s (2009) diagnostics for markedness: Number of inventories in 584 585 P-base containing the indicated segments 100 . . . . . . . t k kʷ 22 / 42

Surprising numbers TYPOLOGICAL SURPRISES Applying Clements’s (2009) diagnostics for markedness: Number of inventories in 584 585 P-base containing the indicated segments 100 6 . . . . . . . . . t k kʷ tʷ 22 / 42

69.5 33.1 24.6 . . . . . . t tʷ kʷ 24.5 . . . k TYPOLOGICAL SURPRISES Surprising numbers Mean numbers of consonants in inventories in P-base containing the indicated segments 23 / 42

69.5 33.1 24.6 . . . . . . t tʷ kʷ Surprising numbers TYPOLOGICAL SURPRISES Mean numbers of consonants in inventories in P-base containing the indicated segments 24.5 . . . k 23 / 42

69.5 33.1 . . . . tʷ kʷ TYPOLOGICAL SURPRISES Surprising numbers Mean numbers of consonants in inventories in P-base containing the indicated segments 24.5 24.6 . . . . . t k 23 / 42

69.5 . . tʷ TYPOLOGICAL SURPRISES Surprising numbers Mean numbers of consonants in inventories in P-base containing the indicated segments 33.1 24.5 24.6 . . . . . . . t k kʷ 23 / 42

TYPOLOGICAL SURPRISES Surprising numbers 69.5 Mean numbers of consonants in inventories in P-base containing the indicated segments 33.1 24.5 24.6 . . . . . . . . . t tʷ k kʷ 23 / 42

…but /t/ is not any more marked than /k/. e markedness of complex segments is not simply the sum of their features. TYPOLOGICAL SURPRISES Surprising numbers By these tests, /tʷ/ is much more marked than /kʷ/… 24 / 42

e markedness of complex segments is not simply the sum of their features. TYPOLOGICAL SURPRISES Surprising numbers By these tests, /tʷ/ is much more marked than /kʷ/… …but /t/ is not any more marked than /k/. 24 / 42

TYPOLOGICAL SURPRISES Surprising numbers By these tests, /tʷ/ is much more marked than /kʷ/… …but /t/ is not any more marked than /k/. e markedness of complex segments is not simply the sum of their features. 24 / 42

2 Why do we find contrastive rounding primarily on velars, then on uvulars and labials, and only rarely on coronals? TYPOLOGICAL SURPRISES Two puzzles e typological paern presents two puzzles: 1 Why is contrastive rounding on consonants so oen deployed in an apparently uneconomical way? 25 / 42

TYPOLOGICAL SURPRISES Two puzzles e typological paern presents two puzzles: 1 Why is contrastive rounding on consonants so oen deployed in an apparently uneconomical way? 2 Why do we find contrastive rounding primarily on velars, then on uvulars and labials, and only rarely on coronals? 25 / 42

. PROPOSAL . PROPOSAL 26 / 42

Peripheral is in opposition to Coronal. (Cf. Jakobson, Fant & Halle’s (1955) opposition grave vs. acute.) Basic place representations look something like this (seing aside some aspects of underspecification): p t k ʔ Peripheral Coronal Peripheral Labial Dorsal PROPOSAL Place features Rice & Avery (1993); Rice (1995, 2002): Dorsal and labial places are encompassed by Peripheral. 27 / 42

(Cf. Jakobson, Fant & Halle’s (1955) opposition grave vs. acute.) Basic place representations look something like this (seing aside some aspects of underspecification): p t k ʔ Peripheral Coronal Peripheral Labial Dorsal PROPOSAL Place features Rice & Avery (1993); Rice (1995, 2002): Dorsal and labial places are encompassed by Peripheral. Peripheral is in opposition to Coronal. 27 / 42

Basic place representations look something like this (seing aside some aspects of underspecification): p t k ʔ Peripheral Coronal Peripheral Labial Dorsal PROPOSAL Place features Rice & Avery (1993); Rice (1995, 2002): Dorsal and labial places are encompassed by Peripheral. Peripheral is in opposition to Coronal. (Cf. Jakobson, Fant & Halle’s (1955) opposition grave vs. acute.) 27 / 42

PROPOSAL Place features Rice & Avery (1993); Rice (1995, 2002): Dorsal and labial places are encompassed by Peripheral. Peripheral is in opposition to Coronal. (Cf. Jakobson, Fant & Halle’s (1955) opposition grave vs. acute.) Basic place representations look something like this (seing aside some aspects of underspecification): p t k ʔ Peripheral Coronal Peripheral Labial Dorsal 27 / 42

But what if it can also represent /kʷ/? In other words, labialized velars (in some systems) are represented as a distinct major place of articulation. e fact that one of the two places of articulation is phonetically secondary might not need to be explicit in the phonological representations. PROPOSAL Place features is is an obvious representation for a labial-velar: k͡p Peripheral ▼ qqqqqqq ▼ ▼ ▼ ▼ ▼ ▼ Dorsal Labial 28 / 42

In other words, labialized velars (in some systems) are represented as a distinct major place of articulation. e fact that one of the two places of articulation is phonetically secondary might not need to be explicit in the phonological representations. PROPOSAL Place features is is an obvious representation for a labial-velar: k͡p Peripheral ▼ qqqqqqq ▼ ▼ ▼ ▼ ▼ ▼ Dorsal Labial But what if it can also represent /kʷ/? 28 / 42

e fact that one of the two places of articulation is phonetically secondary might not need to be explicit in the phonological representations. PROPOSAL Place features is is an obvious representation for a labial-velar: k͡p Peripheral ▼ qqqqqqq ▼ ▼ ▼ ▼ ▼ ▼ Dorsal Labial But what if it can also represent /kʷ/? In other words, labialized velars (in some systems) are represented as a distinct major place of articulation. 28 / 42