Abstract

Drawing on a two-million-word corpus of Sanskrit, the article documents and analyzes two previously unrecognized generalizations concerning the morphoprosodic conditioning of retroflex spreading (nati). Both reveal harmony to be attenuated across the left boundaries of roots (i.e., between a prefix and a root or between members of a compound), in the sense that while harmony applies across these boundaries, when it does so, it accesses a proper subset of the targets otherwise accessible. This attenuation is analyzed here through the “ganging up” of phonotactics and output-output correspondence in serial Harmonic Grammar. The article also simplifies the core analysis of the spreading rule, primarily through recognizing FLAPOUT, an articulatorily grounded constraint.

Sanskrit exhibits a consonant harmony process called nati by which retroflexion spreads progressively and at any distance from a retroflex continuant trigger to a coronal nasal target (e.g., (1a–b)), assuming that no consonantal coronal intervenes to block it (1c). A trigger can occupy any morphological position, including a prefix (1d).

(1)

  • a.

    √ɻaːɡʱaυ-eːna → [ɻaːɡʱaυ-eːɳa] ‘by the descendant of Raghu’

  • b.

    √ɻuɡ-na- → [ɻuɡ-ɳa-] ‘broken’

  • c.

    √ɻatʰ-eːna → [ɻatʰ-eːna] ‘by the chariot’

  • d.

    pɻa-√ɦi-noː-ti → [pɻa-ɦi-ɳoː-ti] ‘incites’

Nati has drawn the attention of linguists for nearly three thousand years. Among generative phonologists, it has played significant roles in treatments of harmony, (non)iterativity, feature geometry, autosegmentalism, and prosodic phonology (section 1), and it continues to inform new developments. Recently, for instance, Jardine (2014) identified nati as one of only two known segmental (as opposed to tonal) processes in the world’s languages with the potential to be “unbounded circumambient,” that is, sensitive to unbounded contexts on both sides of the target (see section 4). Hansson (2010:189–191) identifies several respects in which nati is unusual among consonant harmony systems, including the nonoverlap between triggers and target, the coronal blocking of a coronal harmony, the progressive directionality, and the (occasional) phrasal domain. One might add that prefixes rarely initiate harmony crosslinguistically (Baković 2000, Hyman 2002, Krämer 2003, Kenstowicz 2009).

The present article has two goals. First, it simplifies previous analyses of the core facts of nati, primarily through incorporating into the analysis a phonetic property of retroflex stops, namely, flapping out (i.e., releasing in a more anterior position). Sanskrit is argued to be normal typologically in that its retroflex stops flap out, while its retroflex continuants do not. This constraint explains a number of seemingly disparate properties of nati, including its trigger set, its noniterativity, its progressive directionality, and some aspects of blocking (section 2).

Second, drawing on a two-million-word corpus of Vedic and Epic Sanskrit, this article revisits the primary data, identifying and analyzing two previously unrecognized (including by the grammars) morphological conditions on nati. Both independently reveal the left boundaries of roots to attenuate spreading, in the sense that harmony accesses fewer targets after it crosses a boundary. In particular, cross-boundary harmony never affects immediately postplosive targets, whereas stem-internal harmony almost always does so (section 3). Moreover, cross-boundary harmony rarely accesses targets in preretroflex position, whereas stem-internal harmony always does so (section 4).

Both cases are analyzed in Harmonic Grammar (HG) through the “ganging up” of the relevant independently motivated markedness constraint (*Tɳ or the Obligatory Contour Principle) with an output-output correspondence constraint, IDENTOO([retro]), which requires derived forms to match their bases’ retroflexion. As a brief illustration of this principle, stem-internal harmony almost always accesses postplosive targets (e.g., [√ɻuɡ-ɳa-] ‘broken’, [√tɻp-ɳu-] ‘be pleased with’), revealing that a proharmony constraint—say, SHARE—outweighs *Tɳ ( plus input-output IDENTIO). Harmony also normally applies across root boundaries (e.g., [pɻa-√hi-ɳu-] ‘incite’, [paɻj- √aŋk-aːɳ-aːm] ‘of the beds’); thus, SHARE > (IDENTOO + IDENTIO). But when both of these situations arise simultaneously, as when harmony must cross a root boundary to reach a postplosive target, harmony fails (e.g., [pɻa-√bʱaɡ-na-] ‘crushed’, [pɻ-√aːp-nu-] ‘attain’). This generalization is captured if the summed weight of (*Tɳ + IDENTOO + IDENTIO) exceeds that of SHARE.

This analysis is argued to be superior to other conceivable approaches not involving HG, serialism, or output-output correspondence (section 5). Optimality Theory (OT) approaches relying on morphological indexation or constraint conjunction are critiqued in sections 5.1–5.2. A stratal OT account is addressed in section 5.3. Finally, nonserial HG, while able to capture the gang effects described here, is arguably more pathological than its serial counterpart (section 5.4).

1 The Language and the Corpus

The basic facts surrounding nati ([nəti]; English pronunciation ['nʌti]) have been recounted numerous times since antiquity. Pāṇini (ca. 500–350 BCE) treats them in a set of 39 rules in the last chapter of the last book of his grammar, the Aṣṭādhyāyī (8.4.1–39; see Böhtlingk 1887:461–472, Vasu 1898:1651–1670). Nati is also discussed in the Prātiśākhyas, ancient treatises on Vedic pronunciation (Wackernagel 1896:188, Allen 1951:940).

The term nati, literally ‘bending, curvature’ (Allen 1953:66), was not used by Pāṇini; rather, it appears in the Prātiśākhyas (

graphic
k-Pr. 5.61, Vājasaneyi-Pr. 1.42). It refers to tongue retroflexion as an articulatory process. To refer to the retroflexes as a class, the ancient phoneticians used a different term, mūrdhanya ‘cerebral’, from mūrdhán, the relevant passive articulator. The term nati sometimes appears in print erroneously with an initial retroflex, but is properly dental-initial, as in the Prātiśākhyas, coming from a zero-grade nominalization (< *n
graphic
-ti) of the root √nam ‘bend’, the same root found in ‘namaste’ (nam-as=te), literally ‘[a] bow [to] thee’. It does not, as the spelling *ɳati would suggest, mean something like ‘ɳ-ification’. While nati can in principle refer to any process of retroflexion, it is used here, as elsewhere, to refer only to retroflex harmony affecting nasals.

Notable modern grammatical descriptions include those by Whitney (1889:64–66), Brugmann (1897:352, 849), Macdonell (1910:38–40), Allen (1951:940–946), Renou (1952:55–58, 1961:16–18), and Wackernagel and Debrunner (1957:102–107). In terms of the coverage of the data, however, Wackernagel (1896) is hardly superseded by these or other works (cf. Grammont 1950:251–252, Collinge 1965, Langendoen 1968:84, Burrow 1973:97, among others).

Among generative works, nati has featured prominently in analyses of consonant harmony, feature geometry, autosegmental spreading, and prosodic phonology, including those by Johnson (1972:13–61), Vergnaud and Halle (1978), Selkirk (1980:122–125), Kiparsky (1985:113), Sagey (1986:134), Schein and Steriade (1986:717–719, 720–723), Steriade (1986, 1995), Avery and Rice (1989:192–193), Cho (1991), Shaw (1991), Rice and Avery (1991), Kaun (1993), Clements and Hume (1995:289), Flemming (1995a:112–113, 1995b), Humbert (1995:192–205), Ní Chiosáin and Padgett (1997:35–41), Gafos (1999:207–214, 220–224), Halle, Vaux, and Wolfe (2000: 423–424), Hansson (2001:223–243, 2010:179–193), Hamann (2003:122–123, 195–196), Rose and Walker (2004:518–519, 2011:284–285), Kaplan (2008:20–21), Graf (2010:71–76), Jurgec (2011:20–24), Arsenault (2012:144–150), Cathcart (2012:79ff.), Jardine (2014:15–16), and others. Even in 1951, Allen could already refer to nati as “only too well-known” ( p. 940). Half a century later, Gafos (1999:177, 209) could identify it as both “notorious” and “a prototypical case of long-distance assimilation,” though most analyses, including those by Allen (1951) and Gafos (1999), analyze it as strictly local spreading. These various strains of research are cited as relevant below.

The language names used in this article, while standard, deserve comment, since different authors employ them with different degrees of specificity. First, Sanskrit here refers to all of Old Indic (also known as Old Indo-Aryan). It is not used here to refer only to Classical Sanskrit, as it sometimes is elsewhere. Sanskrit in this broad sense can in turn be divided at the coarsest into two periods, the older Vedic (ca. 1500–600 BCE) and the younger Classical (ca. 600– BCE), the latter more closely conforming to Pāṇini’s rules (Masica 1993:50–55). Classical thus construed subsumes the two Sanskrit epics.

When this article cites corpus counts, they derive from the texts enumerated in table 1, all downloaded from the Göttingen Register of Electronic Texts in Indian Languages.1 The texts, arranged roughly by chronology (the

graphic
g-Veda being the oldest extant Sanskrit text), are labeled according to period and genre. Abbreviations are given in parentheses. For example, 10v 1b 5e would mean that the form is attested 16 times in the corpus: 10 times in the Vedas, 1 time in the Brāhmaṇas, and 5 times in the epics. The corpus includes over two million words in total, roughly one-third Vedic and two-thirds Epic. Since the corpus is not exhaustive, additional forms from dictionaries, grammars, and other texts are also cited when relevant, though not included in corpus statistics.

Table 1
Sources and abbreviations covered in the corpus reports in the text. Each is given with its period, genre, and orthographic word count.
PeriodGenreTextWord count
Vedic Vedas (v
graphic
g-Veda 
164,767 
   Sāma-Veda 19,019 
   Atharva-Veda 85,021 
 Brāhmaṇas (b(Mādhyaṃdina) Śatapatha 127,255 
   Pañcaviṃśa 42,700 
   Gopatha 31,267 
   (Bāṣkala) Kauṣītaki 39,060 
 Early Upaniṣads (uBr
graphic
hadāraṇyaka 
16,502 
   Chāndogya 13,968 
Epic (e  Mahābhārata 1,258,457 
   Rāmāyaṇa 213,773 
    Total: 2,011,789 
PeriodGenreTextWord count
Vedic Vedas (v
graphic
g-Veda 
164,767 
   Sāma-Veda 19,019 
   Atharva-Veda 85,021 
 Brāhmaṇas (b(Mādhyaṃdina) Śatapatha 127,255 
   Pañcaviṃśa 42,700 
   Gopatha 31,267 
   (Bāṣkala) Kauṣītaki 39,060 
 Early Upaniṣads (uBr
graphic
hadāraṇyaka 
16,502 
   Chāndogya 13,968 
Epic (e  Mahābhārata 1,258,457 
   Rāmāyaṇa 213,773 
    Total: 2,011,789 

The consonant and vowel inventories of Sanskrit are shown in tables 2 and 3, respectively (e.g., Cardona 2003). While this article employs the IPA for citing data (though not for names of texts, technical terms, etc.), the IPA can be easily converted back to the standard romanization using these tables. IPA transcriptions below depart from these tables only in giving the short low vowel as [a], as it is normally transcribed, despite its schwa-like quality. For the handful of vowels for which the Vedic and Classical values differ, the Classical values can always be assumed, as is standard practice. The letter anusvāra (), usually said to be a kind of placeless but moraic nasal coda (cf. Japanese), is omitted from the table.

Table 2
Sanskrit consonant inventory, with standard Indologists’ transcription in italics followed by IPA. Asterisked phones are specifically Vedic, [ϕ] and [x] being variant pronunciations of . The chart includes phones usually assumed to be allophonic, namely, [ɲ], [ŋ], [h], [ϕ], [x], [Ɩ], and [Ɩʱ].
LabialDentalRetroflexPalatalVelarGlottal
Plosive p [p] t [t]  [ʈ] c [c] k [k]   
 ph [pʰ] th [tʰ] ṭh [ʈʰ] ch [cʰ] kh [kʰ]   
 b [b] d [d]  [ɖ] j [ɟ] g [ɡ]   
 bh [bʱ] dh [dʱ] ḍh [ɖʱ] jh [ɟʱ] gh [ɡ ʱ]   
Fricative  [ϕ]* s [s]  [
graphic
ś [ɕ]  [x]*  [h] 
           h [ɦ] 
Nasal m [m] n [n]  [ɳ] ñ [ɲ]  [ŋ]   
Lateral   Ɩ [l] [Ɩ]*       
     ḷh[Ɩʱ]*       
Rhotic     r [ɻ]       
Glide v [υ]     y [j]     
LabialDentalRetroflexPalatalVelarGlottal
Plosive p [p] t [t]  [ʈ] c [c] k [k]   
 ph [pʰ] th [tʰ] ṭh [ʈʰ] ch [cʰ] kh [kʰ]   
 b [b] d [d]  [ɖ] j [ɟ] g [ɡ]   
 bh [bʱ] dh [dʱ] ḍh [ɖʱ] jh [ɟʱ] gh [ɡ ʱ]   
Fricative  [ϕ]* s [s]  [
graphic
ś [ɕ]  [x]*  [h] 
           h [ɦ] 
Nasal m [m] n [n]  [ɳ] ñ [ɲ]  [ŋ]   
Lateral   Ɩ [l] [Ɩ]*       
     ḷh[Ɩʱ]*       
Rhotic     r [ɻ]       
Glide v [υ]     y [j]     
Table 3
Sanskrit vowel and syllabic consonant inventory. As before, asterisked transcriptions are Vedic pronunciations. All items can be considered phonemic.
FrontCentralBack
High i [i]   u [u] 
 ī [iː]   ū [uː] 
Mid e [eː] ([ai]*)   o [oː] ([au]*) 
Low   a [ə]   
   ā [aː]   
Diphthong ai [ai] ([aːi]*)   au [au] ([aːu]*) 
Syllabic C 
graphic
 
[ɻ]     
 
graphic
 
[ɻː]     
 
graphic
 
[ḷ]     
FrontCentralBack
High i [i]   u [u] 
 ī [iː]   ū [uː] 
Mid e [eː] ([ai]*)   o [oː] ([au]*) 
Low   a [ə]   
   ā [aː]   
Diphthong ai [ai] ([aːi]*)   au [au] ([aːu]*) 
Syllabic C 
graphic
 
[ɻ]     
 
graphic
 
[ɻː]     
 
graphic
 
[ḷ]     

The rhotic, a retroflex continuant and by far the most common trigger of nati, is transcribed here, with its syllabic variants, as [ɻ], though it may have been (or varied with) tapped or trilled [ɽ]. Whitney (1889:secs. 24, 52–53), for one, identifies it as untrilled, noting, among other things, that “[n]o authority hints at a vibration as belonging to it,” as might be expected for a trill, given the general articulatory detail commanded by the ancient phoneticians. Indeed, one ancient prescription refers to excessive contact (atisparśa) as a barbarism (barbaratā) (Allen 1951:54). Other possible but not strong hints at the smoothness of the rhotic include its productive participation in s-rhotacism (Catford 2001), its frequent metatheses and glide-like alternations in syllabicity, and its status as a reflex of both *r and *l (Catford 2001). Furthermore, as section 2.2 elaborates, the fact that the rhotic initiates a domain of progressive retroflex spreading indicates that it does not “flap out” into a more anterior position on its release. Since retroflex stops and flaps typically flap out, while retroflex fricatives do not, this diagnostic might also support a smooth rhotic, though a tap/trill is not ruled out.2

As the heading of table 2 implies, dental /n/ vs. retroflex /ɳ/ is a phonemic contrast in Sanskrit (e.g., [paːna] ‘drinking’ vs. [paːɳa] ‘stake in a game’). Nevertheless, its functional load is low, the vast majority (over 80%) of tokens of [ɳ] being due to nati.3

This article uses final hyphens in citing words only when they would be hyphenated in standard romanization. The lack of a hyphen does not imply that the word could stand alone as such. For example, the word nati itself could never occur as nati without an ending (e.g., nominative singular [nati-h]), but is normally cited as nati, not nati- or natiḥ. Internal hyphens, which are often problematic, are supplied freely when convenient, but always when the morphology is relevant to the application of nati. As is also common practice in citing Sanskrit words, pitch accent is marked when convenient (generally when a word is being quoted from a text in which accent is marked), though lack of a marked accent does not imply that the word lacks an accent or that its location is unknown.

2 Triggers, Targets, Blockers, and the Importance of Flapping Out in Their Analysis

2.1 Preliminary Data

Nati is a progressive (left-to-right) consonant harmony. Its triggers are all and only the nonlateral retroflex continuants,{ɻ

graphic
graphic
ː
graphic
} (on the status of {Ɩ Ɩʱ} as (non)triggers, which has not previously been discussed, see section 2.4). Its lone target is the dental /n/, which becomes retroflex [ɳ]. Harmony obtains across an arbitrarily long string of segments so long as no blocker intervenes. Blockers (also called opaque segments) comprise the consonantal (i.e., excluding [j]) coronals. These basic properties are summarized in (2). The domain is typically the word (though occasionally larger or smaller). For the most part (though see section 3), harmony is blind to morphology. For example, a rhotic in a prefix will target a visible nasal in a root, suffix, infix, or other prefix; a rhotic in a suffix will target subsequent suffixes; and so forth. Syllabic position is also irrelevant. Nati applies only if the target immediately precedes a vowel, glide, or nasal; on this restriction, see section 2.4.

(2)

graphic

As an illustration, consider the instrumental singular suffix /-eːna/ (see also, e.g., Hansson 2010:179–185 for a different presentation of the basic data). When attached to a stem lacking a trigger, it surfaces as such, as in (3). (The vs. 0 addendum to a corpus citation makes it explicit that no counterexample is found in the corpus; in general, however, patterns suggested by example sets are entirely regular unless otherwise noted.) When the stem contains an (unblocked) trigger, the suffix undergoes nati, as shown in (4).

(3)

graphic

(4)

graphic

As mentioned, harmony is blocked by an intervening coronal. This subsumes the dental, retroflex, and palatal series, with the one exception of the palatal glide /j/, which is always transparent (as in (4b,g)). Some blockers are exemplified in (5). Items (5e–f ) also reinforce that retroflex stops do not serve as triggers (see also (3c–d)).

(5)

graphic

Certain coronals, while possible to analyze as blockers, cannot be illustrated in blocking position. First, /ḷ/ and /ɟʱ/, while expected to block, are rare and unattested in diagnostic positions in the corpus. Lacking evidence to the contrary, they are assumed to behave like /l/ and /ɟ/. The situation is similar for the palatal nasal, which is only attested adjacent to a palatal stop in the corpus and therefore cannot be isolated as a blocker, though it is presumed to be one. Second, as previously observed (Gafos 1999:213, Arsenault 2012:147), the triggers—all coronal—are ambiguous in their status as blockers, since they could be either transparent or blocking with retriggering; see (6). The status of these segments as blockers is therefore free to follow from theory-internal considerations.

(6)

graphic

Finally, the dental nasal cannot occur in blocking position because it itself undergoes harmony, becoming [ɳ]. In such cases (as with underlying /ɳ/, which is not a trigger), harmony does not spread beyond the undergoing /n/ to the next /n/; see (7). Thus, coronal nasals can also be considered blockers.

(7)

graphic

2.2 Core Analysis

The facts introduced to this point are analyzed in this section; additional complications will be considered in sections 3 and 4. A key and often overlooked component of their explanation, it is maintained here, concerns flapping out (Ladefoged 1964), a property of retroflex stops (including nasals) by which the tongue tip moves forward during the closure phase of the segment, releasing into a more anterior position (Ladefoged 1964, Bhat 1973:47, Dave 1977, Dart 1991, Shalev, Ladefoged, and Bhaskararao 1993, Butcher 1995, Krull et al. 1995, Steriade 1995:5–6, Spajić, Ladefoged, and Bhaskararao 1996, Dart and Nihalani 1999, Simonsen, Moen, and Cowen 2000, Flemming 2003, Hamann 2003, Boersma and Hamann 2005, Arsenault 2012). Retroflex stops are therefore contour segments, so to speak, and could be narrowly transcribed as such: for example, narrow [

graphic
] for broad [ʈ] (Boersma and Hamann 2005:21–24). The narrower transcription does not imply that the release of a retroflex stop is homophonous with that of a dental stop; it indicates only that the release enters an anterior (e.g., alveolar) configuration (Steriade 1995:6).

The acoustic consequence of flapping out is that the F3 depression associated with retroflexion is realized more prominently in the VC than the CV transition. Flapping out has been documented palatographically for retroflex stops of Australia, Scandinavia, and South Asia (including daughters of Sanskrit such as Hindi and Gujarati; Boersma and Hamann 2005:18) and is further corroborated by their phonological behavior, particularly their better cueing by left-hand context (e.g., Steriade 1995, Hamann 2003). But flapping out does not apply to all retroflexes. As Boersma and Hamann (2005:18) clarify, while it is a typical, perhaps even universal, property of retroflex stops, it appears not to characterize retroflex fricatives (see also Bhat 1973:47 and Flemming 2003:346 in tentative agreement with this caveat). The lack of flapping out of retroflex fricatives is also supported by their phonology, particularly their frequent interactions with following vowels (Boersma and Hamann 2005:18–19).

It is therefore assumed on both typological and internal grounds that the Sanskrit retroflex stops flap out, while the retroflex fricative does not. Internal grounds include the behavior of stops vs. fricatives in nati, as explained presently, as well as their licensing requirements: retroflexion is contrastive for stops only in postvocalic position (with marginal exceptions due to onomatopoeia and dialect borrowing), while the retroflex fricative is more broadly distributed (e.g., [

graphic
áʈ] ‘six’ vs. [sát] ‘being’). Thus, [
graphic
] is narrowly [
graphic
], not [
graphic
]. Note that retroflex continuants also possess stronger internal cues to their anteriority, which could also support their broader licensing.

While the typology is less clear for retroflex rhotics, internal grounds support treating Sanskrit [ɻ] like [

graphic
] in terms of flapping out. Aside from its comparably broad licensing (e.g., it occurs word-initially, where it remains retroflex, as confirmed by nati), the fact that both [ɻ] and [
graphic
] serve as triggers for progressive retroflexion is itself prima facie evidence of their lack of flapping out, given that the consensus holds nati to be a spreading harmony (e.g., Flemming 1995b, Gafos 1999, Ní Chiosáin and Padgett 2001, Rose and Walker 2004, Hansson 2010, Jurgec 2011). That the mechanism of nati is strictly local spreading (i.e., gestural extension) as opposed to agreement across nonundergoing interveners is supported by the existence of blockers, progressive directionality, disjoint triggers and target, and the (occasional) phrasal domain (see the above-cited works, especially Hansson 2010:189–191). In order to initiate a progressive domain of retroflex spreading, the retroflex continuants cannot flap out. This asymmetry between stops and continuants is summarized in (8).4 In what follows, retroflex stops will continue to be given their broad transcriptions, with the understanding that they flap out.

(8)

graphic

The constraint enforcing flapping out in stops is here called FLAPOUT. Loosely speaking, this constraint requires every retroflex coronal stop to have an anterior offset. Coronal is specified because noncoronal stops can link to [retroflex] on this analysis (as when retroflexion spreads through them), and noncoronal retroflexes such as [ḳ] are not accompanied by flapping out. In terms of autosegmental spans, the constraint, as in (9), demands that every retroflex coronal stop coincide with the right edge of its span of retroflexion.

(9) FLAPOUT: Penalize every retroflex coronal stop that is nonfinal in its span of retroflexion.

One other caveat is that only released retroflex stops flap out. A cluster such as /ɳʈ/, for instance, is presumably realized as a single coronal gesture [ɳ

graphic
] rather than as [
graphic
graphic
]. The latter, which contains a dental stop between two retroflex stops, can be ruled out by other constraints (much as, say, [kqk] would be). GEN may also produce candidates in which such clusters share their [retroflex] feature. In such candidates, the no-line-crossing convention and ( possibly GEN-encoded) NOGAP, which forbids discontinuous spans (Kiparsky 1981, Archangeli and Pulleyblank 1994, Walker 2014), together ensure that the first part does not flap out.

Next, a constraint is required to motivate the harmony, whose mechanism appears to be strictly local spreading as opposed to long-distance agreement with intervening nonundergoers, as discussed. Several constraint-based approaches to spreading can be found in the literature, including ALIGN, SPREAD, SPECIFY, *A-SPAN, AGREE, and ∀-HARMONY; see Wilson 2003 and McCarthy 2009a, 2011 for overviews and pathologies of these proposals. Here, SHARE([retro]) is employed, following McCarthy’s (2009a, 2011) schema, as in (10).

(10) SHARE([retro]) (abbreviated SHARE): For every pair of adjacent segments, assign a penalty if they are not both linked to the same token of [retroflex].

Given the autosegmental setting, the spreading feature is often taken to be privative, as with [retroflex] here, agreeing with recent analyses of Sanskrit (e.g., Ní Chiosáin and Padgett 2001) and other languages (e.g., McCarthy 2009a, 2011, Walker 2014). This assumption is not crucial here; if binary [anterior] or [TTCO] (tongue tip constriction orientation; Gafos 1999) were employed instead, the constraint definitions could be recalibrated. Also following McCarthy (2009a, 2011), SHARE([retro]) is taken to be violated by a pair of adjacent segments in which neither segment is linked to [retroflex].

A competing faithfulness constraint, IDENT([retro]) (11), penalizes changing a segment’s anteriority. In the tableaux, this constraint is taken to be violated by /n/ → [ɳ] even though the latter, assuming it flaps out, retains an anterior release. At any rate, since IDENT is not an active constraint here, this detail of formulation is irrelevant. For a fuller analysis of retroflex licensing and contrast in Sanskrit, see section 2.3.

(11) IDENT([retro]) (abbreviated IDENT): Penalize a segment whose anteriority differs from that of its input correspondent.

The constraint-based framework employed here, for reasons to be clarified in section 5, is serial Harmonic Grammar (Kimper 2011, Mullin 2011, Pater 2012), which is the same as Harmonic Serialism (HS; McCarthy 2009a, 2011) except set in Harmonic Grammar (HG; Legendre, Miyata, and Smolensky 1990, Smolensky and Legendre 2006, Pater 2009b, Potts et al. 2010) rather than Optimality Theory (OT; McCarthy and Prince 1993, Prince and Smolensky 1993). Serial HG is like classical OT and HG in that each language comes with a fixed ranking or weighting of constraints. Unlike in classical OT and HG, however, only one operation (e.g., addition or deletion of an association line) is permitted per GEN/EVAL cycle, and the output of each evaluation is recycled as an input to a new evaluation until no more changes are optimizing, at which point the derivation converges. Furthermore, since it is serial HG rather than OT, constraints have real-valued nonnegative weights and the violation score of a candidate is the weighted sum of its violations, which are taken to be nonpositive integers. The candidate with the greatest harmony wins. On harmony in HS, as well as more general background on the theory, see McCarthy 2009b, 2011 and references therein.

Tableau series (12) illustrates the derivation of [ɻaɳa] ‘delight’ from ( possible input) /ɻana/. Parentheses indicate spans of retroflexion, that is, strings in which every segment is linked to the same token of [retroflex]. Retroflexion is redundantly marked on every segment within the span, using an underdot if the IPA lacks a symbol. Since only one operation is permitted per step, the span grows one segment at a time until it reaches target /n/, at which point it cannot spread any further without violating higher-weighted FLAPOUT. Reducing or deleting the span is never optimizing. In Step 1, candidate (c), which removes the retroflex span altogether by anteriorizing the rhotic, violates SHARE three times, one for each pair of adjacent segments, following McCarthy’s definition and use of that constraint. When the most faithful candidate wins, as in Step 3, the derivation converges. As is also standard in HS, input-output correspondence constraints such as IDENT are evaluated with respect to the input to the current step, not the original input.

(12)

graphic

The simple weighting of FLAPOUT > SHARE, while not yet the full story, already captures several core features of nati. First, it captures the stop-continuant asymmetry in triggering without specifying it in the harmony apparatus, as reinforced by (13) with [muːɖʱ-eːna] ‘by the fool’. Because retroflex stops flap out (not only in Sanskrit, but perhaps universally), they cannot trigger. As (13) also illustrates, this analysis predicts regressive retroflexion insofar as no blocker interferes (blocking is treated below). No harm comes from this prediction, for two reasons. First, if it were incorrect, one could add a constraint preventing leftward spreading such as INITIAL([feat]) (McCarthy 2004, 2009a:9). But the prediction is not incorrect, at least not on language-internal grounds. Sanskrit orthography distinguishes retroflexion only in coronals. It follows that retroflexion in noncoronals is effectively hidden structure (granting also the impossibility of instrumental study) and free to follow from analytical and typological considerations (Allen 1951:940–946, Steriade 1995:51).

(13)

graphic

Second, the analysis predicts the directionality of nati without specifying it in the harmony apparatus. Consider /υaːnaɻa/ ‘monkey’ in (14). Retroflexion spreads onto the vowels surrounding /ɻ/, but cannot affect the preceding /n/, given that [ɳ] continued by retroflexion would violate FLAPOUT. Thus, the system embodies the prediction that retroflex spreading harmony targeting stops could only possibly be progressive, as in Sanskrit. Regressive retroflex spreading harmony is attested, as in Kinyarwanda (Walker and Mpiranya 2005, Walker, Byrd, and Mpiranya 2008), but its targets are continuants, not stops, consistent with this proposal. This proposal also does not make any predictions about retroflex harmony by correspondence as opposed to spreading (see Arsenault 2012). When multiple orders of operations are tied, only one path is illustrated.

(14)

graphic
graphic

Third, as (15) illustrates (for [k

graphic
aɳ-eːna] ‘by the instant’), the analysis captures the fact that harmony terminates when it reaches a target, rather than continuing on to yet another target. In other words, the noniterativity of harmony is derived from an independent property of the language rather than implemented as an ad hoc parameter or constraint.5

(15)

graphic

Fourth, harmony is asymmetric in the sense that an anterior continuant does not cause an unblocked retroflex nasal to become anterior (e.g., /sa-ɡaɳa/→ [sa-ɡaɳa], *[sa-ɡana] ‘along with troops’). This follows from the statement of SHARE, which favors the spreading of retroflexion, but not of anteriority. While the present ranking predicts *[

graphic
aɡaɳa] for this input, the prevention of segments such as /s/ from undergoing harmony is treated in (18). The point here is that anterior continuants are not triggers like retroflex ones.

Fifth, and finally, FLAPOUT covers blocking by retroflex stops (e.g., [υiɻaːʈ-eːna] ‘by Virāṭa’). Retroflex continuants (e.g., [k

graphic
iːɻ-eːɳa] ‘by milk’) are also handled appropriately, since the retroflex span is free to spread to /n/ regardless of the multiplicity of triggers. This leaves only blocking by palatals (except /j/) and dentals. The former can be motivated biokinematically (and hence potentially by GEN), in that a palatal articulation is incompatible (in Sanskrit, if not universally) with tongue tip retroflexion (Gafos 1999:213–214, 223–224; cf. Flemming 2003, Hamann 2003, Boersma and Hamann 2005). As Gafos (1999:214) also emphasizes, this articulatory incompatibility naturally fails to extend to the palatal vocoid, which involves less arching of the tongue body.

At this point, then, FLAPOUT > SHARE remains incomplete concerning the core data only in that (a) it fails to restrict the targets to /n/ as opposed to the other anteriors (viz., /t tʰ d dʱ s l ḷ/) and (b), relatedly, it fails to capture blocking by anteriors, which are thus far predicted to undergo harmony en route to a target just like noncoronals. For example, the correct output for /ɻas-eːna/ is [ɻas-eːna] ‘by flavor’, in which /s/ both blocks and fails to undergo retroflexion. But the ranking so far generates *[ɻa

graphic
-eːɳa], in which /s/ is both transparent to and undergoes retroflexion.

Following Ní Chiosáin and Padgett (1997:36; also Ohala and Ohala 1993, Padgett 1995, Steriade 1995, 2009, Gafos 1999), place is generally less faithful for nasals than for other consonants, particularly obstruents. For one, nasals are more likely to undergo assimilation, all else being equal. Moreover, diachronically, a contrast between dental and retroflex is less robust for nasals than for plosives, as suggested by the daughters of Sanskrit that lost the /n ~ ɳ/ contrast while preserving phonemic retroflexion in the plosives (e.g., Bengali, Nepali, Hindi dialects; Masica 1993). A solution, then, is to rank SHARE below a faithfulness constraint that prevents retroflexion from spreading onto oral coronals, for example, IDENTORCOR in (16). This general strategy of FAITH[specific] ≫ HARMONY ≫ FAITH[general] is not new here; it is employed by all prior constraint-based analyses of nati (see below) to implement the asymmetry between /n/ and other dentals.

(16)

graphic

In essence, while this approach assumes that [retroflex] can link to any segment (except perhaps the palatals), its interaction with coronals, especially oral coronals, is afforded special faithfulness owing to its greater perceptibility on them. Tongue tip orientation during noncoronals is less tightly regulated. Gafos (1999:222) employs FAITH(TTCO, Obstruent) to this end, but this constraint fails to account for blocking by /l/ and for the transparency of noncoronal obstruents. The analysis of Ní Chiosáin and Padgett (1997:36) is dispersion/contrast-based, evaluating paradigms as candidates (see Flemming 1995a). The approach here is more classical and predicts blocking to be independent of the contrastive status of retroflexion in coronals. In Sanskrit, after all, all coronals block, but retroflexion is contrastive for only a subset of them. In particular, there is no anteriority contrast in the laterals in any period, but laterals block in all periods (as do the palatals, for which retroflexion is moot). While one could still maintain that laterals block because retroflexion is contrastive for some coronals in Sanskrit, or because retroflexion is a possible contrast for laterals typologically, invoking contrast at all is unnecessary. The greater perceptibility of retroflexion differences in coronals, especially oral coronals, can be projected onto faithfulness constraints (cf. Steriade 2009). While this explanation still invokes dispersion in some sense, it does not require evaluating paradigms as candidates.

Derivation (17) illustrates both blocking of harmony by an oral coronal and failure of the same oral coronal to undergo harmony. Derivation (18) shows that anticipatory harmony to a coronal continuant is also properly ruled out.

(17)

graphic

(18)

graphic

To summarize thus far, the ranking for basic nati, including its trigger and target sets, directionality, noniterativity, retroflex-anterior asymmetry, and transparent vs. blocking segments, is depicted as a Hasse diagram in (19).

(19)

graphic

This analysis improves upon previous constraint-based analyses of nati (full OT analyses being offered in Ní Chiosáin and Padgett 1997, 2001 and Gafos 1999; see also sketches in Steriade 1995 and Jurgec 2011). First, the proposed proharmony constraint is the simplest, merely stating the feature that spreads. The constraint says nothing about the set of triggers, targets, or directionality; all of these properties fall out from interaction with other relatively simple and independently motivated constraints. Compare the proharmony constraints in (20)–(23), all of which include one or more features of the triggers and/or target (viz., continuancy and/or coronality), as well as directionality.

(20) TIP POSITION: “A nasal apical maintains the same tip position, raised or lowered, as a preceding continuant apical.” (Steriade 1995:51)

(21) ALIGN-R([retro], C): “Align any [retroflex] feature contained in a [+continuant] segment Sm to a consonant Sn, where n > m.” (Ní Chiosáin and Padgett 1997:36)

(22) HARMONY(TTCO = [retroflex], trigger = [+continuant]) [in which TTCO refers to tongue tip constriction orientation and HARMONY is defined essentially as ALIGN-R] (Gafos 1999:218–223)

(23) ALIGN-R(phonological-phrase, [−anterior], [+coronal]) ( Jurgec 2011:23)

In the present analysis, the interaction of SHARE and FLAPOUT captures several seemingly disparate properties of nati, including the restriction of triggers to continuants, the progressive directionality (given that the target is a stop), the blocking by retroflex stops, and the noniterativity of spreading, in the sense that harmony cannot spread through the first eligible target to any following target.

The celebrated noniterativity of nati (see, e.g., Kiparsky 1985:113, Gafos 1999:213, Kaplan 2008:21, Hansson 2010:190) is here an artifact of the target being a stop. In other constraintbased analyses, noniterativity is stipulated or left unanalyzed (see also footnote 5 on a common rule-based approach). It is stipulated through a dedicated, rankable constraint in Jurgec 2011:23 and through alignment directly to the target (as opposed to a domain edge) in Ní Chiosáin and Padgett 1997:36. As Hansson (2010:186–188) explains, the analysis proposed by Gafos (1999) fails to account for noniterativity, and alignment-to-target analyses fail to properly handle blocking, at least given the vague formulation of target selection in (21). The analysis here not only covers noniterativity, but also requires it of spreading-driven retroflex harmonies targeting stops. A hypothetical version of Sanskrit with the same phonetics except for iterative nati, or nati feeding another progressive retroflexion, could not exist. At the same time, if the target of retroflexion is a continuant, nontermination and feeding are predicted. Indeed, this prediction is borne out by Sanskrit. Consider ruki, another rule of progressive retroflexion, by which a rhotic, velar, or nonlow vowel causes immediately following /s/ to become retroflex (Selkirk 1980, Beguš 2012). Ruki, as predicted, invariably feeds nati, as in /υ

graphic
graphic
-ana/ → [υ
graphic
graphic
-aɳa] ‘sprinkling’, in which the rhotic first triggers retroflexion in the sibilant, which in turn triggers retroflexion in the nasal (recall that [s] would otherwise block nati). This is possible because the target of ruki retroflexion is a continuant, unlike the target of nati retroflexion.

While this section has treated the basic properties of nati, including its triggers, targets, blockers, directionality, and noniterativity, all of which are known in the phonological literature, some additional complications are documented and analyzed in sections 3–4.

2.3 Addendum concerning Contrast

Retroflexion is contrastive only among coronals in Sanskrit. On the present approach, noncoronals can also bear the feature in Sanskrit, but only noncontrastively, to accommodate harmony or assimilation. In this respect, Sanskrit differs from a language such as Badaga, in which retroflexion is contrastive on vowels (Emeneau 1939 et seq.). A richness-of-the-base input such as /ạ/ in Sanskrit must therefore neutralize to [a] and also fail to trigger nati.

As an illustration, consider the richness-of-the-base inputs /(ɻ)a-na/ and /p(ạ)-na/ for desired outputs [(ɻạ-ɳ)a] and [pa-na], respectively. To be clear, [(

graphic
ạ-ɳ)a] would not be ill-formed per se, but failure to suppress nati in such cases would erroneously permit the existence of coronalfree morphemes that trigger nati, such as a prefix /pạ-/, perhaps written pa, but triggering nati on a following root. No such morpheme exists.6

Assuming that retroflex noncoronals (including vowels) are marked, *RETRO-NC penalizes them (NC for noncoronal, possibly split up into multiple constraints). *RETRO-NC > IDENT([retro]) causes the neutralization of /ạ/ and /a/ to [a], as in (24). Candidate (d), in which /ạ/ triggers nati, loses because *RETRO-NC and IDENTCor([retro]) (“A coronal must retain its input specification for retroflexion”) collectively outweigh SHARE. Implicit in this analysis is a P-map ( perceptual map) or *MAP hierarchy of faithfulness constraints (Zuraw 2007, 2013, McCarthy 2009b, Steriade 2009): IDENTOrCor([retro]) > IDENTCor([retro]) > IDENT([retro]), expressing the fact that changes in retroflexion are most perceptible on oral coronals, followed by nasal coronals and then noncoronals (see section 2.2). Convergence steps and ([retro]) are omitted from tableaux in this section to save space.

(24)

graphic

Meanwhile, in the context of harmony, spreading across vowels and other noncoronals remains optimal because SHARE outweighs *RETRO-NC and IDENT combined, as in (25).

(25)

graphic

Furthermore, as (26) illustrates, underlying /ɳ/ is properly preserved as such.

(26)

graphic

A final technicality concerns the possibility of multisegment spans of retroflexion preexisting in the input. For example, what if the prefix were not merely /p(ạ)/, as in (24), but /(

graphic
ạ)/? The analysis as it stands predicts that /(
graphic
ạ)/ should trigger nati, as in (27). In other words, a prefix like /(
graphic
ạ)/ is indistinguishable from one like /(ɻạ)/. This treatment is incorrect if it is assumed that a coronal-free morpheme cannot trigger nati.

(27)

graphic

At least two antidotes are available. First, multisegment span structure could be absent from lexical representations (as tentatively entertained in McCarthy 2004:5), which might either lack span structure altogether (much as inputs are often assumed to lack prosodic structure; see McCarthy 2008:303) or limit it to single segments. In either case, a prefix with the underlying form /(

graphic
ạ)/ could not exist. Second, assuming headed spans (McCarthy 2004 et seq.), a constraint could penalize a span of retroflexion with a noncoronal (or non-oral-coronal) head (e.g., *DEPENDENT-HEAD in Mullin 2011).

Although *RETRO-NC and IDENTCor were not made explicit in the tableaux in section 2.2, their inclusion in those derivations with the present weights does not alter any of the outcomes shown there.

2.4 Addenda concerning the Basic Rule

Two details concerning the basic rule are yet to be addressed. First, the triggers for nati are usually reported to be the retroflex continuants, which include {ɻ

graphic
graphic
ː
graphic
}. But the Vedic inventory, as table 2 suggests, also includes laterals [Ɩ] and [Ɩʱ], presumably also retroflex continuants. They appear exclusively as allophones of /ɖ/ and / ɖʱ/, respectively, in intervocalic position in certain Vedic texts. Judging by 45 diagnostic tokens in the present corpus, retroflex laterals never trigger nati.

Possible causes for this failure include the following. First, it could be synchronic opacity, with lateralization counterfeeding nati. Second, it could be that the apparent opacity is not synchronic but a historical artifact. Under this scenario, at the time of composition, the stops would have been pronounced as stops. At some later point in the transmission of the text, lateralization would occur, but without retriggering nati, either because nati had lost productivity, or because the nasals’ anteriority was orthoepically fixed. Finally, it is possible that the class of triggers was synchronically not the retroflex continuants, but the central retroflex continuants, potentially with phonetic motivation. Given the laterals’ shallow origin in stops, for instance, perhaps they continued to flap out in articulation, in which case they could not trigger. In any case, given the rarity of these allophones and the irrelevance of this issue to the remainder of this article, these questions are left open.

Second, nati is usually reported to apply only if the target immediately precedes a vowel, glide, or nasal, that is, a nonliquid sonorant. As Schein and Steriade (1986:720–722) motivate (also Hansson 2010:183), failure before a liquid, fricative, or word boundary follows from general phonotactics independent of nati, such as word-final neutralization. Only nonapplication before a plosive (e.g., /caɻ-a-n-ti/ → [caɻ-a-n-ti] ‘wander (3pl.)’) requires further comment, as retroflex nasal-plosive clusters are otherwise permitted (e.g., /phaɳ-ta/ → [pʰaɳ-ʈa] ‘spring ( pass. part.)’). On the present approach, the step at which the intermediate form [c(ạɻạ)nti] would yield [c(ạɻạɳ)ti] needs to be precluded. This can be accomplished with an appropriate version of CODACOND (McCarthy 2008) penalizing heterorganic nasal-plosive clusters. If CODACOND is given a weight of, say, 5, the derivation converges at [c(ạɻạn)ti], as desired: the collective violation of CODACOND, IDENTCor, and IDENT outweighs the benefit of spreading to /n/. /pʰaɳ-ta/ → [pʰaɳ-ʈa] is also handled appropriately with this addition. Retraction, as in [(

graphic
hạ)ɳ-ta], would cost two more violations of SHARE than expansion, as in [(
graphic
hʰạɳ-ʈ)a]. The latter violates IDENTORCor (both candidates violate IDENTORCor, IDENT, and CODACOND equally), but not enough to outweigh its benefit from SHARE. As discussed in section 2.2, homorganic nasal-stop clusters, being single stop gestures, do not violate FLAPOUT.

3 Boundary Attenuation I: Postplosive Targets

One aspect of nati often omitted from generative discussions is that while velars and labials are normally transparent, as illustrated in (4), they often block when immediately preceding the target nasal. For example, consider the verb stem [pɻ-√aːp-] ‘attain’ (from preverb [pɻa] + root √aːp). Nati applies without exception whenever the target nasal is postvocalic, as in (28) and numerous similar examples. But when the nasal immediately follows the final [p] of the stem, as in (29), nati always fails. This failure is not, moreover, a function of the [nu]/[noː] suffix (class 5 present stem formative), as (30) illustrates using the same preverb and suffix but a vowel-final root.

(28)

graphic

(29)

graphic

(30)

graphic

The postplosive blocking of nati in (29) no doubt reflects a more general phonotactic of Sanskrit. While /n/ and /ɳ/ generally contrast (section 1), the contrast is virtually confined to tautomorphemic postvocalic (occasionally postsonorant) position (Steriade 1995). Retroflex nasals can be found in postplosive position, but only as a result of assimilation. Putting aside nati contexts, if the plosive is coronal, the following coronal nasal must agree in place (e.g., [ɻátna] ‘gift’, [aʈɳaːɻá] ( proper name), [jaɟɲá] ‘sacrifice’); otherwise, the coronal nasal must be dental (e.g., [sυápna] ‘sleep’, [aɡní] ‘fire’). No isolated lexemes like *[sυapɳa] or *[aɡɳi] are found. Thus, Tɳ appears to be more marked than Tn.

The analysis from section 2.2 can be easily amended by adding a highly ranked constraint forbidding postplosive retroflexes, such as *Tɳ in (31). While this constraint could likely be generalized further—for example, to palatal and velar (but not labial) nasals—these details of formulation are unimportant here. Retroflex plosive-nasal clusters (e.g., [aʈɳaːɻá]) can be motivated by assimilatory constraints dominating *Tɳ, not shown. If *Tɳ + IDENTCor _ IDENT + SHARE, as in (32), postplosive nati is suppressed.

(31) *Tɳ : Penalize a retroflex nasal immediately following a plosive.

(32)

graphic

Weighting (32) is incorrect, however, since nati does regularly target a postplosive nasal in some forms. The data in (33) cover all such forms in the corpus (see section 1), sorted by descending frequency. Irrelevant affixation and compounding are now factored out in the entries, such that only the relevant root and affix, if any, are shown. For example, (33d) [ɻéːkɳas] ‘inheritance’ includes counts for [ɻéːkɳas] in various case forms as well as prefixed [su-√ɻéːkɳ-aːh] ‘well-endowed (masc. nom. sg.)’ and suffixed [√ɻéːkɳas-υatiː] ‘endowed (fem. nom. sg.)’. The vs. 0 annotation indicates that the lexeme never occurs in the corpus as *[ɻéːknas], regardless of genre, period, or morphological context.

(33)

graphic

By contrast, all of the forms in the corpus in which an otherwise eligible postplosive /n/ fails to undergo nati are given in (34).7 When the trigger is not explicitly shown, as in (34c), assume that the ‘X-’ portion contains a visible trigger. For example, (34c) (√)X-√ɡʱna ‘X-killer’ subsumes [√n

graphic
-√ɡʱná] ‘man-killer’, [√υ
graphic
tɻa-√ɡʱná] ‘V
graphic
tra-killer’, and so forth, generalizing over irrelevant affixation and compounding as before. Similarly, when ‘preverb-’ is indicated in the gloss, all applicable trigger-containing preverbs (e.g., [pɻa-]) are included.

(34)

graphic

The difference between (33), in which nati applies to postplosive targets, and (34), in which it does not, is that in all of the cases in (33), no initial root boundary intervenes between trigger and target, whereas in almost all of the cases in (34) (with a handful of exceptions to be addressed below), an initial root boundary intervenes. This root boundary criterion separates tokens into the two categories with almost perfect accuracy (100% hits and no misses for the first set; >99% hits and <1% misses for the second). Furthermore, it holds across genres and periods. On its lack of recognition in the previous literature, see the end of this section.

To be sure, some of the nati failures in (34) could be attributed to compounds failing to undergo nati by virtue of being compounds. In Classical Sanskrit, after all, nati often fails to apply across compound boundaries. In Vedic, however, in which nati usually applies across compound boudaries, it never does so when the target is postplosive. Consider, for example, two derivatives of the root √ɦan ‘kill’, namely, /ɡʱná / ‘killer’ and /ɦána/ ‘killing’, in compound-final position. When the first member of the compound contains a trigger, /ɦána/ undergoes nati, while /ɡʱná/ does not; see (35) and (36), respectively.

(35)

graphic

(36)

graphic

In any case, compounds are not the whole story. Even preverbs that otherwise normally trigger nati in their stems never affect a postplosive target. This was already demonstrated in (28)–(30); some additional examples involving /pɻa-/ are given in (37) and (38). Other trigger-containing prefixes (e.g., [paɻi-], [du

graphic
-]) behave the same.

(37)

graphic

(38)

graphic

To summarize thus far, first, noncoronal plosives are normally transparent to nati, as established in section 2.1 and reinforced here. Coronals, for their part, always block. A noncoronal plosive also blocks if and only if (a) it immediately precedes the target and (b) the trigger and target straddle a root boundary. The latter configuration is found both when the trigger occupies a prefix and when the trigger occupies a preceding member of a compound.

Schematically, the new generalization can be summarized as in (39), where √ notates the left edge of a root. As the organization of (39) implies, nati failure in (39c) can be analyzed by the “ganging up” (e.g., Jäger and Rosenbach 2006, Kenstowicz 2009, Pater 2009b:1008ff.) of the two markedness constraints implied by (39a) and (39b) against SHARE. In other words, while a violation of neither (39a) alone nor (39b) alone is enough to prevent nati, when both (39a) and (39b) are violated, nati fails in just this “worst-of-the-worst” case scenario.

(39)

  • a.

    Harmony is marked across √.

  • b.

    Retroflexion is marked immediately following a plosive.

  • c.

    Nati applies in spite of (39a) and (39b), except when both apply simultaneously.

Formally, ganging up can be analyzed using weighted constraints, as in HG (section 2.2). This situation obtains when the weights of two weaker constraints sum to a value greater than that of the stronger constraint (i.e., w1 < w3; w2 < w3; w1 < w2 < w3). In the present case, the stronger constraint is SHARE, and one of the two weaker constraints is *Tɳ. The other must penalize cross-√ harmony. The approach adopted here to do so (see section 5 regarding other possibilities) is output-output correspondence (e.g., Benua 1995, 1997, Kenstowicz 1996, Ussishkin 1999, Steriade 2000, McCarthy 2005, Zuraw 2013), in particular, IDENTOO([retro]) in (40). The base of correspondence of a prefixed form is its unprefixed counterpart (as with Italian in Kenstowicz 1996).8 Members of compounds also stand in correspondence with their uncompounded bases. For example, [√ɦi-noː-ti] is the free base corresponding with prefixed [(

graphic
ɻạ- √
graphic
ị-ɳ)oː-ti] ‘incites’. [(
graphic
ɻạ- √
graphic
ị-ɳ)oː-ti] therefore incurs three violations of IDENTOO, one for each segment that undergoes retroflexion.

(40) IDENTOO ([retro]) (abbreviated IDENTOO): Assign a penalty for every segment that differs in anteriority from its correspondent in the base.

Derivation (41) shows a prefix triggering nati when *Tɳ is not at stake. Because SHARE outweighs IDENTOO (plus the other IDENT constraints), harmonizing across √ is optimal. The convergence step, in which harmony stops at [ɳ] owing to FLAPOUT (section 2.2), is omitted.

(41)

graphic

Derivation (42) shows nati accessing a postplosive target when √ is not crossed. Because /√ɻeːknas/ (assuming a richness-of-the-base input without retroflexion) is a root, IDENTOO is not applicable. Derivation (42) ignores the debuccalization of final /s/ to [h] that would occur if this word were pronounced in isolation.

(42)

graphic
graphic

Finally, when the span crosses √ and reaches a potential postplosive target, as in [pɻ-√aːpnoː-ti] in (43), both *Tɳ and IDENTOO are violated, now collectively (with IDENT and IDENTCOR) outweighing SHARE. This gang effect prevents nati from reaching the target in Step 4. Vowel coalescence between preverb /pɻa/ and root /aːp/ is assumed by fiat.

(43)

graphic

This section will now conclude with some remarks on exceptions and on the lack of previous recognition of the rule described here. The following representative passage from a grammar says only this about postplosive targets (Wackernagel 1896 says somewhat more, but the outlook for the present point is the same):

The immediate combination of n with a preceding guttural or labial seems in some cases to hinder the conversion to

graphic
: thus, vṛtraghn
graphic
etc., kṣubhnāti, tṛpnoti (but in Veda tṛp
graphic
u), kṣepnú , suṣumná. (Whitney 1889:sec. 195a)

This description implies that postplosive targets vary freely, as indeed phonologists mentioning this caveat have taken it (Steriade 1995:52–53, Hansson 2010:182). It is the nature of grammars, afterall, to list exceptions without tempering them with clear indications as to the robustness of the rule. To address Whitney’s examples: [√υ

graphic
tɻa-√ɡʱnáː] ‘V
graphic
tra-killer’ follows the rule proposed here. [√k
graphic
ubʱ-nV] ‘shake’ does not, but it is entirely absent (with either [n] or [ɳ]) from the present two-million-word corpus. [√t
graphic
ρ-NV] ‘be pleased’, in which N ∈ {n, ɳ}, occurs eight times and breaks the rule only once. In other words, Whitney foregrounds the exception, not the rule. [√k
graphic
eːp-nú] ‘springing’ occurs once and is a genuine exception. [su-√
graphic
um-nV] ‘gracious’ is not included in the lists above, which consider only postplosive targets. Its counts here are ‘v5’ for [n] and ‘v1 b1 e7’ for [ɳ].9

[su-√

graphic
um-nV] and two similar forms from the list of nonundergoers in (34)—namely, [υi- √
graphic
kabʱ-na] ‘fix’ and [d(a)u(h)-√
graphic
υápn-ja] ‘bad sleep’—require further comment. In all three, the trigger [
graphic
] ostensibly occupies the root, and none exhibits nati. While at first glance exceptions to the proposed generalization, in fact they follow from it. In every case, the trigger acquires its retroflexion from the prefix via ruki (section 2.2). Thus, they correspond to nonprefixed forms without nati, and the gang effect of *Tɳ or *Nɳ (footnote 9) with IDENTOO applies in the prefixed forms, properly suppressing nati.

Putting aside these three forms with ruki as explained, then, the rule, as stated above, is a near-perfect generalization. All 138 tokens with postplosive nati have a domain of retroflexion that is root-initiated, and 1,648 of 1,650 (99.9%) of tokens with a failure of postplosive nati have a domain of retroflexion that would have to cross √. The only robust exceptions in this corpus are 1 token of [√k

graphic
eːp-nV-] and 1 of [√t
graphic
p-nV-] (against 7 of [√t
graphic
p-ɳV-]), both mentioned by Whitney (1889).

4 Boundary Attenuation II: Clashing Spans

As a further complication, nati also fails under certain predictable circumstances when a retroflex follows the target. For example, consider once again the preverb [pɻa-], now with the root √naɕ- ‘vanish’ (or ‘reach’). As section 3 demonstrated, [pɻa-] triggers nati in a root or suffix. /pɻa- √naɕ-/ is no exception, as (44) reinforces

(44)

graphic

But when the final consonant of √naɕ - is realized as retroflex (owing to irrelevant morphophonology), nati fails in the vast majority of instances, as shown in (45). Forms (45b–c) do not appear in the present corpus, but are cited as such in the sources given. In total, the corpus contains 320 instances of /pɻa-√naɕ-/. Of them, 218 have a nonretroflex ending, and nati applies in 100% of those cases. The remaining 102 have a retroflex ending, and nati fails in 91% of those cases.

(45)

graphic

Moreover, the pattern suggested by the above paradigm is general. Regardless of the prefix and root involved, nati fails to cross √ when an unblocked retroflex follows the target (unblocked meaning that no coronal intervenes; cf. [t] in [pɻa-√ɳeː-t

graphic
] ‘leader’). Additional examples are given in (46). Diagnostic forms are infrequent because the requisite setup is quite specific, being a triggering prefix attached to a stem with an unblocked target followed by an unblocked retroflex, which in Sanskrit is unlikely to be provided by a suffix. But insofar as forms meeting these criteria are found, the generalization is supported.

(46)

graphic

Aside from the 9 exceptional (against 91 regular) tokens of [pɻa-√ɳa

graphic
-ʈa-] mentioned in (45), the only other cases in the corpus in which a prefix triggers nati in a root domain containing an unblocked retroflex are given in (47).10 These exceptions are discussed further at the end of this section.

(47)

graphic

That nati is suppressed by a following retroflex is already established in the literature (Macdonell 1910:sec. 47, Allen 1951:945–946, Hansson 2010:184), though to my knowledge no formal analysis of this suppression has been put forth. Hansson (2010:184) suggests that it might arise from misperception, specifically, the hypocorrective misattribution of the source of the cues for retroflexion on the nasal to the surrounding retroflexes. But given the data to be presented in (48), this explanation cannot be correct: in other contexts, Sanskrit orthoepy/orthography consistently records retroflexion on nasals in interretroflex position.

Previous discussions do not make explicit the fact that suppression is confined to a limited morphological context. Hansson (2010:184), for one, reports only that “[w]hen there is also an /

graphic
/ or /r/ later in the word, retroflexion fails to apply” (likewise Graf 2010, Jardine 2014).11 Indeed, Macdonell (1910), whom Hansson cites, leaves this interpretation open. Allen (1951) claims that the suppressing retroflex may be at most one vowel away from the target (on which, see below), but omits any mention of morphological conditioning.

In particular, suppression of nati by a following retroflex occurs only when the span must cross √. Otherwise, nati applies regardless of whether a retroflex follows, as the examples in (48), among numerous others, illustrate. Such cases of nonsuppression within the root-suffix complex vastly outnumber the cases of cross-√ suppression considered above.

(48)

graphic

Descriptively, the new generalization can be summarized as in (49), whose structure mirrors that of (39). As (49) implies, a gang effect with IDENTOO is once again in evidence.

(49)

  • a.

    Harmony is marked across √.

  • b.

    Retroflexion is marked immediately preceding another domain of retroflexion.

  • c.

    Nati applies in spite of (49a) and (49b), except when both apply simultaneously.

What remains to be treated is the markedness constraint implied by (49b). Here, it is proposed that the failure of harmony in such cases reflects the OCP (Obligatory Contour Principle; Leben 1973, McCarthy 1986, Myers 1997). OCP([retro]) in (50) penalizes every point of contact between two spans of retroflexion. On this approach, [pɻa-√nɻt], for instance, fails to undergo nati because doing so would give *[(pɻạ-√ɳ)(

graphic
)t], which violates both the OCP and IDENTOO. (Other constraints in section 2, namely, FLAPOUT, IDENTORCOR, and MAX([retro]), prevent fusing or deleting the autosegments.)

(50) OCP([retro]) (abbreviated OCP): Penalize adjacent domains of retroflexion.12

The analysis of suppression then runs as follows. First, the fact that a prefix such as [pɻa-] normally triggers harmony across √ continues to hold, as (51) illustrates.

(51)

graphic

Second, when no √ intervenes, an OCP violation is tolerated, as in (52). Once again, candidates with fusion and deletion are ruled out by other constraints in section 2.13

(52)

graphic
graphic

Finally, when √ would interrupt harmony, harmony fails, thanks to the OCP and the IDENT constraints collectively outweighing SHARE. In (53) Step 1, IDENTOO ensures that retroflexion in the root spreads first. As candidate (c) of Step 4 demonstrates, the second span of retroflexion cannot retreat across the vowel to rescue the OCP. Step 4 does not contain a candidate that involves both a retraction of retroflexion before [

graphic
] and an application of nati to /n/, since only one change per step is possible. IDENTOO thus does double duty in this derivation, first by favoring root spreading over prefix spreading, precluding an ultimate outcome as *[(
graphic
ɻạ√ɳ)a(
graphic
ʈ)a-], and second by contributing to the gang effect with the OCP.

(53)

graphic

The OCP([retro]) analysis correctly predicts that nati should not be suppressed if the following span of retroflexion is one or more segments removed from the preceding one. In [(pɻạ-√ɳ)eː-t(

graphic
)], for instance, the second span does not spread across /t/ (on oral coronal blocking, see section 2.2). Therefore, the OCP is not applicable, and nati succeeds.

Beyond harmony, the activity of OCP([retro]) in Sanskrit is corroborated by reduplication. Consider the desiderative, which comprises a CV reduplicant prefix in which V is high as well as the suffix /-s/ (Whitney 1889:secs. 1026–1040). If the root is /s/-initial and no retroflex follows, the root undergoes ruki conditioned by the prefix, as in [si-√

graphic
aː-s-] ‘wish to gain’ (for √saː) and [su-√
graphic
up-s-] ‘wish to sleep’ (for √sυap). But if the suffix undergoes ruki, ruki in the root is usually (though not always) suppressed, as in [si-√saŋk-
graphic
-] ‘wish to hang’ (for √saɲɟ) and [si-siːɻ-
graphic
-] ‘wish to flow’ (for √s
graphic
).

This section will now conclude with some remarks on the locality of nati suppression. First, all examples of suppression so far have involved a target in a root. Nevertheless, some of the forms used to exemplify suppression in previous research (Macdonell 1910:sec. 47, Allen 1951: 945–946, Hansson 2010:184) have a somewhat different profile. These are enumerated in (54), in which ‘=’ indicates a compound boundary and ‘X’ contains an unblocked trigger.

(54)

graphic

In all of the cases in (54), the suppressed target is in a prefix. The compounds in (54a–b) are handled properly by this analysis if it is assumed that each member is evaluated separately by IDENTOO (e.g., for (54a), the bases would be [puɻú] and [ni

graphic
graphic
ídʱ-] with its prefix).14 Items (54c–d) are cases of double prefixation in which the first prefix could trigger nati in the second, but fails, perhaps because a retroflex follows. However, these prefix pairs are likely ineligible for nati in the first place.15

Second, one might wonder whether suppressing retroflexes are confined to roots, given that the examples of nonsuppression mostly involve inflectional suffixes. This is not the case, as properly captured by the analysis. Prefixes suppress in (54); the suffix [-

graphic
] suppresses in (45c); and in the other two cases in (45), the suppressing consonant, though ostensibly located in the root, acquires its retroflexion from a suffix.

Third, all examples of (successful) suppression in this section involve a morpheme-initial target. The analysis here predicts that suppression should also be possible for non-morphemeinitial targets. Forms confirming or disconfirming this prediction are rare, owing to the combined infrequencies of the requisite parts, namely, a triggering prefix, a root beginning with a vowel or noncoronal followed by a dental nasal, and a suffix containing an unblocked retroflex. One such case was cited in (47), namely, [pɻa-√υaɳ-éː-

graphic
u]. Suppression fails in this case, but this single token is not particularly compelling, since other causes for the failure are conceivable (e.g., next paragraph).

A final question concerns whether the suppressing retroflex can be any distance from the target. All of the examples of suppression in this section involve a suppressor that is at most

graphic
graphic
C
graphic
away from the target, where
graphic
is a short vowel. No case of suppression across a long vowel or multiple syllables was found. What’s more, all of the exceptions to suppression in (47) involve a target and a suppressor straddling a long vowel. Since diagnostic data are sparse, it will be left unresolved here whether this generalization is principled or accidental. If it were principled, it could indicate that leftward spreading from retroflexes (across noncoronals) is not unchecked, as implied by section 2.2, but rather limited to a single vocalic mora. On the present approach, a constraint requiring a preceding vowel to license retroflexion (cf. RETRO/V ___ in Steriade 1995) could dominate a constraint forbidding leftward spreading (e.g., INITIAL([feat]) in McCarthy 2009a) to achieve minimal leftward spreading.

Suppression also fails regardless of distance when the would-be suppressor is derived by external sandhi, apparently a case of postlexical opacity. For example, /áːɻuɳi-s/ ‘

graphic
runi’ is realized as [áːɻuɳiɻ] when followed closely by a voiced-initial word, as in [áːɻuɳiɻ aːɦa] ‘
graphic
ruṇi said’. While more philological work remains to be done on suppression, this article is not the place to undertake it. This section has shown that nati suppression, previously unanalyzed, can be accommodated by the proposed account of nati by adding a single constraint, OCP([retro]).

5 Analytical Comparisons

The analysis of nati proposed in this article relies on output-output (OO) correspondence in serial Harmonic Grammar (HG). This section considers four alternative approaches (namely, morphological indexation, constraint conjunction, level ordering, and nonserial HG), each with the potential to obviate one or more of these mechanisms (OO, serialism, and/or HG). It is argued that all four alternatives are untenable or pathological.

5.1 Morphologically Indexed Constraints

As established in sections 3–4, harmony affects a postplosive or preretroflex target unless it has to cross a left root boundary (√) to reach it. In the proposed analysis, IDENTOO tempers the benefit of spreading into the root domain. But consider an alternative approach in OT by which prefixes are treated as weak triggers for harmony, in the sense that they access fewer targets than root triggers do. Root control is common in harmony systems (Clements 1980 et seq.), and affixtriggered harmony can be penalized directly (e.g., Kenstowicz 2009). Nevertheless, the nati data in sections 3–4 cannot be characterized as affix weakness: a root trigger is comparably weak when its span crosses √, as seen in compounds (e.g., (35), (36), (54)). Thus, the descriptive generalization is not that affixes are weak triggers, but that spreading is weakened by the left edges of roots.16

Rather than treating affixes as weak, a related OT strategy might treat roots or stems as strong. Assume that stem here refers to the root-suffix complex. Generic SPREAD([retro]), as defined in (55), penalizes unharmonized segments regardless of their morphological affiliation. SPREAD ([retro])Stem (56) penalizes unharmonized segments within a stem when the head of the span occupies the same stem.17

(55) SPREAD ([retro]): For every [retroflex] token a1, penalize every segment to which a1 is not associated.

(56) SPREAD ([retro])Stem: For every [retroflex] token a1 whose head is in stem b1, penalize every segment in b1 to which a1 is not associated.

SPREAD ([retro]) Stem ≫ *Tɳ ≫ SPREAD([retro]) handles the morphological conditioning properly, as in (57) and (58), in which heads are underlined. In (57), stem-initiated harmony is strong, compelling violation of *Tɳ. In (58), by contrast, harmony initiated from outside of the stem is weak, and *Tɳ decides.

(57)

graphic

(58)

graphic

This analysis covers the core boundary-attenuation facts discussed in this article without invoking serial HG or OO correspondence. Nevertheless, it has drawbacks unshared by the HG analysis, including too-many-solutions pathologies. First, since stem indexation relies on headed spans (footnote 17), an alternative to spreading within the stem is shifting the head outside of the stem. In (59), candidate (d) should win (as candidate (d) does in (57); see also section 3), but candidate (a), with head shifting and weak spreading, prevails.

(59)

graphic

This situation could be patched in Sanskrit by ranking a constraint requiring head position faithfulness above SPREADStem (e.g., FTHHDSP in McCarthy 2004). But even then, it remains as an unwanted typological prediction, given the factorial typology. More generally, SPREAD/ALIGN constraints in nonserial OT exhibit a number of pathologies, some of which are solved by serialism (see Wilson 2003, McCarthy 2004, 2009a, 2011, Kimper 2011).

A further empirical problem for SPREADStem, as an anonymous reviewer observes, concerns forms such as [υi-√

graphic
kabʱ-na-] ‘prop’ (from root √skabʱ), in which [
graphic
] occupies the stem but fails to trigger strong spreading to postplosive /n/ (section 3). The OO-based account properly handles such cases without any additions, since the unprefixed base is [√skabʱ-na-], without retroflexion, and IDENTOO therefore gangs with *Tɳ. For SPREADStem to handle such cases, the head of retroflexion would have to escape to the prefix, but that behavior is undesirable in other contexts, as illustrated by (59).

A more general problem for morphological indexation is that the domain has to be stipulated. Why is “Stem” the root plus suffixes, as opposed to, say, the root plus prefixes, or just the root, or the root plus certain suffixes, and so forth? Are these other constraints in the constraint set? The OO-based account is more restrictive and arguably more explanatory. Whenever a free base (footnote 8) is available, it exerts analogical force via correspondence. In the case of [υi-√

graphic
kabʱ-na-], otherwise identical forms without the prefix are available (√skabʱ-na-), so IDENTOO (ganging with other IDENT constraints) precludes nati. No morphological domain has to be stipulated as part of any constraint.

5.2 Constraint Conjunction

The analyses in sections 34 relied on gang effects and hence on HG. In section 3, for instance, spreading to a postplosive target (in violation of *Tɳ ) is grammatical but only when doing so would not also violate IDENTOO ([retro]). This ostensible gang effect could alternatively be analyzed in OT using constraint conjunction. For this case, IDENTOO &*Tɳ could be fused into a single, hybrid constraint local to some domain (Smolensky 1995). Similarly, for the gang effect in section 4, the conjunction IDENTOO &OCP could be employed.

The domain of these conjunctions cannot be the word: the constraint would then be violated (erroneously) even if the violations of IDENTOO and *Tɳ (or the OCP) came from two unrelated loci of retroflexion in the word, as in compounds.18 It also cannot be the root, since targets often occupy suffixes. One potentially viable domain is the segment, as in (60) and (61). For this domain to work, local to the segment must mean that the segment merely participates in the violating structure, not that it comprises or contains it. *Tɳ , after all, can only be evaluated with respect to pairs of segments. If this looser notion of locality is unacceptable, the domain could still be taken to be the biphone, although the latter is not a phonological constituent.

(60)

graphic

(61)

graphic

At any rate, even putting aside issues of locality, conjunction is known to be highly pathological ( Pater 2009a and references therein). For example, (IDENT(voice)&NOCODA)seg ≫ IDENT (voice) produces a pathological grammar in which voicing is neutralized in onsets but not in codas (Ito and Mester 1998).

Second, viability aside, the conjunction analysis is arguably less elegant than the HG analysis. Consider the respective constraint hierarchies in (62) and (63). Conjunction requires extra, complex constraints to accomplish what is handled by simple constraints alone in HG, albeit with weighting. Moreover, the complex constraints are formally redundant. Both contain IDENTOO, and all four conjuncts have [retro] as a predicate. While these properties make sense for Sanskrit, in terms of pure formalism, they are coincidences. The theory could just as easily encode a language with these conjuncts indexed to unrelated features, such as (IDENTOO ([retro])&OCP([labial]))seg.

(62) (IDENTOO&*Tɳ)seg, (IDENTOO &OCP)seg ≫ SHARE ≫ IDENTOO, *Tɳ, OCP

(63) SHARE > IDENTOO = *Tɳ = OCP

In conclusion, a conjunction analysis may be possible under certain assumptions about locality. It would not obviate the need for OO correspondence (or some suitable replacement), but it would provide an OT alternative to ganging in HG. However, even putting aside concerns about locality and typology, conjunction requires a more complex and redundant constraint set than serial HG does in this case.

5.3 Stratal Optimality Theory

Third, consider a cyclical version of OT that interleaves phonological evaluation with affixation and compounding. This approach is untenable if the ranking is fixed across cycles, since it cannot implement the patterns analyzed as gang effects in sections 34 (see section 5.2). However, stratal OT (Kiparsky 2000, Bermúdez-Otero to appear), in which levels can have different rankings, is more promising. Assume two levels, Stem and Word, such that prefixes are integrated in the Word. In the Stem, SHARE ≫ *Tɳ; therefore, nati affects a postplosive target. In the Word, the ranking is reversed, *Tɳ ≫ SHARE, such that a newly introduced trigger can no longer access a postplosive target.

As it stands, this ranking erroneously undoes the retroflexion of a postplosive target that underwent nati in the Stem. Take /√ɻeːknas/. In the Stem, it becomes [√ɻeːkɳas]. But in the Word, nati is undone by now-dominant *Tɳ, leaving *[√ɻeːknas]. Adding MAXLINK([retro]), which penalizes deleting an association line to [retro] (see Jurgec 2011), solves this problem, preserving [ɳ] that arises in the Stem. MAX([retro]) alone would not work, as the span could retreat without deleting. The rankings are given in (64).

(64)

  • Stem level: MAXLINK([retro]) ≫ SHARE ≫ *Tɳ

  • Word level: MAXLINK ([retro]) ≫ *Tɳ ≫ SHARE

The OCP effects in section 4 would also require more than flipping SHARE and the OCP. Recall from (53) that [(

graphic
ɻạ)-√n(ạ
graphic
-ʈ)a-] wins over *[(
graphic
ɻạ)-√n(ạ
graphic
-ʈ)a-] in the serial analysis because the latter would require two changes in one step. Thus, retroflexion cannot retreat across the vowel to save the OCP. In stratal OT, by contrast, *[(
graphic
ɻạ)-√n(ạ
graphic
-ʈ)a-] is a viable contender. It could be ruled out by adding a constraint—say, LICENSE—that penalizes a retroflex consonant not immediately preceded by a retroflex vowel, as in (65). (LICENSE must be dominated by MAX, DEP, and so on, not shown.)

(65)

  • Stem level: LICENSE ≫ SHARE ≫ OCP

  • Word level: LICENSE ≫ OCP ≫ SHARE

In sum, an analysis of the facts in sections 34 may be possible in stratal OT, though not without additional constraints. The serial HG analysis has the further virtue of requiring only a single, fixed ranking for the language, with arguable benefits for learnability and restrictiveness. Moreover, the stratal analysis requires potentially problematic assumptions about the morphology, since (often highly lexicalized) prefixation and compounding must follow (even inflectional) suffixation. With IDENTOO, a form such as [pɻa√ɦiɳoːti] stands in correspondence with [√ɦinoːti] simply because the former contains the latter and both are free. It does not require prefixation to follow inflection anywhere in the grammar.

5.4 Nonserial Harmonic Grammar

Finally, as mentioned in section 5.1, serialism avoids a number of pathologies exhibited by prospreading constraints in classical OT/HG. Since the analysis here depends on gang effects (section 3), serial HG rather than OT was employed. Serial HG was also favored over classical HG because the latter, while able to implement ganging, is perhaps even more pathological than classical OT when it comes to harmony, predicting what might be called cutoff-point effects (Legendre, Sorace, and Smolensky 2006, Pater, Bhatt, and Potts 2007).

A classical HG cutoff-point pathology is illustrated by the grammar in (66) and (67). In this language, a blocker /ʈ/ is deleted to permit retroflexion to spread further in service of SHARE (already a pathology), but only if more than seven segments would otherwise remain unharmonized. The cutoff need not be only seven; it could be any number, as determined by the ratio of the weight of MAX to that of SHARE. Since harmony is myopic in serial HG, cutoff-point pathologies of this type do not occur ( Pater, Bhatt, and Potts 2007:21).

(66)

graphic

(67)

graphic

6 Conclusion

Sanskrit retroflex spreading is attenuated by left root boundaries, such that stem-internal triggers access more targets than stem-external triggers. At least two independent processes demonstrate this attenuation. First, only stem-internal triggers access postplosive targets. Second, only steminternal triggers access preretroflex targets. These restrictions reveal the activity of two markedness constraints, *Tɳ and the OCP. Permitted to gang with IDENTOO in serial HG, they implement the observed subset relation among triggers. Other possible approaches, including morphological indexation, constraint conjunction, and nonserial HG, were argued to be untenable or pathological. Beyond introducing and analyzing these domain conditions on nati, this article also presented a novel analysis of the basic rule that simplifies previous constraint-based analyses.

Notes

Parts of this article were presented at the 12th Old World Conference in Phonology and the Harvard Indo-European Workshop. I gratefully acknowledge the feedback of those audiences and of Dieter Gunkel, Joe Pater, Rachel Walker, the reviewers, and the editors.

1 gretil.sub.uni-goettingen.de, accessed May 2014.

2 While the rhotic is generally recognized to be retroflex, some ancient phonetic treatises suggest instead that it had an alveolar place (Allen 1951:54–55). As Allen clarifies, even if it were alveolar phonetically (in some dialects), it is clearly functionally retroflex. See Cathcart 2012 on why an anterior rhotic might still induce retroflexion.

3 The present corpus includes 122,680 tokens of [ɳ]. Of these, 82.4% occur in a nati context, though this figure includes occasional false positives in which underlying /ɳ/ happens to occur in a nati context and excludes occasional false negatives in which nati obtains across a word boundary.

4 As mentioned, nontriggering by retroflex laterals is treated in section 2.4.

5 A common refrain of rule-based analyses of nati purports to derive its noniterativity from the fact that a retroflex nasal, the outcome, does not otherwise serve as a trigger, without relating it to any phonetic property (see, e.g., Johnson 1972, Howard 1973, Anderson 1974, Ringen 1976, Kiparsky 1985). On the present analysis, it is no coincidence that the retroflex nasal neither triggers nor propagates, as both are motivated by FLAPOUT. But the present analysis does not relate the (non)iterativity of a harmonic process to whether or not its trigger(s) and target(s) overlap. It predicts a harmony to be possible in which a segment undergoes and propagates the harmony without triggering it. Indeed, if nati is analyzed as strictly local spreading, then this prediction is borne out even by nati: a segment such as [k] undergoes and propagates without being a trigger. As a reviewer notes, other cases of nontriggers propagating harmony can be found in Baiyina Orochen (Kaun 2004) and Seto (Kiparsky and Pajusalu 2003).

6 I am grateful to an anonymous reviewer for raising this issue and for sketching a solution along the lines of the one pursued here.

7 [bɻaɡʱna-] is also found in the corpus (twice in the Sāma-Veda) but omitted from this list since it is a misreading of the Devanāgarī for [bɻaɡʱna-] ‘pale’, in which /n/ is not eligible for nati.

8 Any definition of base selection compatible with this state of affairs is sufficient here. For example, the base of free form i could be defined as the free form j such that j contains the maximal proper subset of the grammatical features of i and no conflicting features (Kager 1999:281).

9 Once plosives and vowels are put aside, only two noncoronals remain that are normally licit in immediately pre-[ɳ] position, namely, [m] and [ɦ]. As [su-√

graphic
um-nV] might suggest, nati applies optionally in this context when nati crosses √. This optionality could be implemented by giving *Nɳ, which penalizes [ɳ] immediately following a sonorant consonant, less weight than *Tɳ in a probabilistic implementation of HG (Hayes and Wilson 2008, Pater 2009b). *Tɳ > *Nɳ can be motivated by the greater perceptibility of retroflexion following a sonorant as opposed to a plosive.

10 Though it is not a verbal form, the Vedic compound [√sυàɻ-√ɳaɻa] ‘sky-man’ (v17 b4 e4 vs. 0) is also an exception to the analysis proposed in this section, perhaps owing to the adjacency of the target and trigger. Other compounds, such as [√ɕik

graphic
aː-√naɻá] ‘trainer, facilitator’ (v3 vs. 0), follow the generalization.

11 The restriction of the following suppressor to retroflex continuants as opposed to retroflex consonants in general is also unmotivated and not assumed here.

12 As an anonymous reviewer points out, flapping out implies that the two retroflex spans in, say, (ɳ)(ạ) are not adjacent under every interpretation. Since this article takes the segment to be the tone-bearing unit for retroflexion, the OCP is interpreted accordingly: for every pair of adjacent segments s1 and s2, if each is linked to a distinct token of [retroflex], assign a violation to OCP. At any rate, even without strict adjacency, autosegmental systems are known to avoid overly rapid excursions. In tonology, *HLH is sometimes invoked to handle such cases (e.g., by McPherson (2016: e52) as “Assign a violation for every sequence HLH in which only a single L association line separates the two H tones”; see also Cahill 2007, Hyman 2010; cf. *TROUGH in Yip 2002). In the present context, the rapid excursion would involve unretroflexing and reretroflexing the tongue tip over the course of a fraction of a segment.

13 For simplicity, the sibilant is given as retroflex in the input, though it is due to ruki. The markedness constraint triggering ruki must be weighted greater than 3.5 in order to outweigh IDENTCOR + IDENT here. If its weight is less than that of SHARE, nati precedes ruki (and subsequent assimilation of the pre-ruki vowel). If its weight is greater than that of SHARE, ruki precedes nati. In either case, the ultimate outcome is the same: the two spans expand until they abut, as illustrated.

14Nati applies less reliably in compounds in general (section 3 and footnote 10), though in this case an argument can still be made for suppression: the initial members of the compounds in (54a–b) comprise [puɻú], [saɦás

graphic
a], [candɻá], and [υaɻ
graphic
á). In every one of the 12 tokens in the
graphic
g-Veda in which one of these initials attaches to a /nV[retro]/ base, nati fails. In every one of the 8 tokens in which one of these initials attaches to any other /n/-initial base, nati succeeds, a significant difference (Fisher’s exact test p < .0001).

15Nati application across certain preverbs is unreliable in Epic Sanskrit. Tellingly, though only 4 tokens of /paɻi-/ before /ni-/ or /nis-/ are attested without a following retroflex, all lack nati: [paɻi-ni-√ɡʱnanta], [paɻi-niɕ-√cit(j)a(m)], and [paɻi-nih-√ɕυasan].

16 A morphological version of *DEPENDENT-HEAD (Mullin 2011) would be untenable here for this same reason, since it defines weakness according to properties of the head of the span (see also Walker 2005).

17 The formulation of SPREAD[(retro)]Stem here assumes headed spans (McCarthy 2004). An alternative formulation might refer only to stem containment: for example, “For every [retroflex] span contained within a stem, penalize every segment within that stem to which the span is not associated.” The problem with this approach is that it predicts root harmony to be suppressible when a prefix is attached. For example, with SPREAD([retro])Stem ≫ *Tɳ, [√(ɻẹːḳɳ)as] wins, as it should, but nati fails if the form is prefixed: for example, *[s(ụ-√ɻẹːḳ)nas]. The prefix provides an “escape hatch” for the span, voiding all of the violations of SPREAD([retro])Stem once the span is no longer stem-contained and thus turning the decision over to *Tɳ. Even if this situation could be remedied with additional constraints in Sanskrit, it would remain as a typological pathology.

18 An example is [pɻa-√ɳaja=√ɡɻaɦ-aɳ=√aːɻth-aːja] ‘for the purpose (artha) of seizing (grahaṇa) affection (?) (praṇaya)’ (e1 vs. 0), in which the OCP violation in grahaṇārthāya is irrelevant for praṇaya.

References

Allen,
W. Sidney
.
1951
.
Some prosodic aspects of retroflexion and aspiration in Sanskrit
.
Bulletin of the School of Oriental and African Studies
13
:
939
946
.
Allen,
W. Sidney
.
1953
.
Phonetics in ancient India
.
Oxford
:
Oxford University Press
.
Anderson,
Stephen
.
1974
.
The organization of phonology
.
New York
:
Academic Press
.
Archangeli,
Diana
, and
Douglas
Pulleyblank
.
1994
.
Grounded phonology
.
Cambridge, MA
:
MIT Press
.
Arsenault,
Paul Edmond
.
2012
.
Retroflex consonant harmony in South Asia
.
Doctoral dissertation, University of Toronto
.
Avery,
Peter
, and
Keren
Rice
.
1989
.
Segment structure and coronal underspecification
.
Phonology
6
:
179
201
.
Baković,
Eric
.
2000
.
Harmony, dominance and control
.
Doctoral dissertation, Rutgers University, New Brunswick, NJ
.
Beguš,
Gašper
.
2012
.
The RUKI-rule in the Rigveda
.
Thesis, University of Ljubljana
.
Benua,
Laura
.
1995
. Identity effects in morphological truncation. In
Papers in Optimality Theory
, ed. by
Jill
Beckman
,
Laura Walsh
Dickey
, and
Suzanne
Urbanczyk
,
77
136
.
University of Massachusetts Occasional Papers in Linguistics 18
.
Amherst
:
University of Massachusetts, Graduate Linguistic Student Association
.
Benua,
Laura
.
1997
.
Transderivational identity: Phonological relations between words
.
Doctoral dissertation, University of Massachusetts, Amherst
.
Bermúdez-Otero,
Ricardo
.
To appear
.
Stratal Optimality Theory
.
Oxford
:
Oxford University Press
.
Bhat,
D. N. Shankara
.
1973
. Retroflexion: An areal feature. In
Working papers on language universals
13
,
27
67
.
Stanford, CA
:
Stanford University
. Available at http://files.eric.ed.gov/fulltext/ED105719.pdf.
Boersma,
Paul
, and
Silke
Hamann
.
2005
.
The violability of backness in retroflex consonants
.
Rutgers Optimality Archive 713
. Available at http://roa.rutgers.edu.
Böhtlingk,
Otto
.
1887
.
Pâṇini’s Grammatik
.
Leipzig
:
H. Haessel
.
Brugmann,
Karl
.
1897
.
Vergleichende Laut-, Stammbildungs-, und Flexionslehre der indogermanischen Sprachen
,
vol. 1
. 2nd ed.
Strassburg
:
Karl J. Trübner
.
Burrow,
Thomas
.
1973
.
The Sanskrit language
.
London
:
Faber and Faber
.
Butcher,
Andrew
.
1995
. The phonetics of neutralisation: The case of Australian coronals. In
Studies in general and English phonetics: Essays in honour of Professor J. D. O’Connor
, ed. by
Jack Windsor
Lewis
,
10
38
.
London
:
Routledge
.
Cahill,
Mike
.
2007
.
More universals of tone
.
Dallas, TX
:
SIL International
.
Cardona,
George
.
2003
. Sanskrit. In
The Indo-Aryan languages
, ed. by
George
Cardona
and
Danesh
Jain
,
104
160
.
London
:
Routledge
.
Catford,
J. C
.
2001
.
On Rs, rhotacism and paleophony
.
Journal of the International Phonetic Association
31
:
171
185
.
Cathcart,
Chundra
.
2012
.
Articulatory variation of the alveolar tap and implications for sound change
. In
UC Berkeley phonology lab annual report
,
76
110
. Available at http://linguistics.berkeley.edu/phonlab/documents/2012/cathcart_phonlabreport.pdf.
Cho,
Young-mee Yu
.
1991
.
On the universality of the coronal articulator
.
Phonetics and Phonology
2
:
159
179
.
Clements,
G. N
.
1980
.
Vowel harmony in nonlinear generative phonology
.
Bloomington
:
Indiana University Linguistics Club
.
Clements,
G. N.
, and
Elizabeth V.
Hume
.
1995
. The internal organization of speech sounds. In
The handbookof phonological theory
, ed. by
John A.
Goldsmith
,
245
306
.
Oxford
:
Blackwell
.
Collinge,
N. E
.
1965
.
Some linguistic paradoxes
.
Journal of Linguistics
1
:
1
12
.
Dart,
Sarah
.
1991
.
Articulatory and acoustic properties of apical and laminal articulations
.
UCLA Working Papers in Phonetics 79
. Available at http://escholarship.org/uc/item/52f5v2x2.
Dart,
Sarah
, and
Paroo
Nihalani
.
1999
.
The articulation of Malayalam coronal stops and nasals
.
Journal of the International Phonetic Association
29
:
129
142
.
Dave,
Radhekant
.
1977
.
Retroflex and dental consonants in Gujarati: A palatographic and acoustic study
.
Annual Report of the Institute of Phonetics, University of Copenhagen
11
:
27
155
.
Emeneau,
Murray B
.
1939
.
The vowels of the Badaga language
.
Language
15
:
43
47
.
Flemming,
Edward
.
1995a
.
Auditory representations in phonology
.
Doctoral dissertation, UCLA, Los Angeles, CA. Published
,
London
:
Routledge (2002)
.
Flemming,
Edward
.
1995b
.
Vowels undergo consonant harmony
.
Paper presented at the Trilateral Phonology Weekend 5, University of California, Berkeley
.
Flemming,
Edward
.
2003
.
The relationship between coronal place and vowel backness
.
Phonology
20
:
335
373
.
Gafos,
Adamantios I
.
1999
.
The articulatory basis of locality in phonology
.
New York
:
Garland
.
Graf,
Thomas
.
2010
.
Logics of phonological reasoning
.
Master’s thesis, UCLA, Los Angeles, CA
.
Grammont,
Maurice
.
1950
.
Traité de phonétique
.
Paris
:
Delgrave
.
Halle,
Morris
,
Bert
Vaux
, and
Andrew
Wolfe
.
2000
.
On feature spreading and the representation of place of articulation
.
Linguistic Inquiry
31
:
387
444
.
Hamann,
Silke
.
2003
.
The phonetics and phonology of retroflexes
.
Doctoral dissertation, Utrecht University
.
Hansson,
Gunnar Ólafur
.
2001
.
Theoretical and typological issues in consonant harmony
.
Doctoral dissertation, University of California, Berkeley
.
Hansson,
Gunnar Ólafur
.
2010
.
Consonant harmony: Long-distance interaction in phonology
.
Berkeley
:
University of California Press
.
Hayes,
Bruce
, and
Colin
Wilson
.
2008
.
A maximum entropy model of phonotactics and phonotactic learning
.
Linguistic Inquiry
39
:
379
440
.
Howard,
Irwin
.
1973
.
A directional theory of rule application in phonology
.
Doctoral dissertation, MIT, Cambridge, MA
.
Humbert,
Helga
.
1995
.
Phonological segments: Their structure and behaviour
.
The Hague
:
Holland Academic Graphics
.
Hyman,
Larry M
.
2002
.
Is there a right-to-left bias in vowel harmony?
Paper presented at the 9th International Phonology Meeting, Vienna
.
Hyman,
Larry M
.
2010
.
Amazonia and the typology of tone systems
. In
UC Berkeley phonology lab annual report
,
376
394
. Available at http://escholarship.org/uc/item/4qr2h33t.
Ito,
Junko
, and
Armin
Mester
.
1998
.
Markedness and word structure: OCP effects in Japanese
.
Ms., University of California, Santa Cruz
.
Jäger,
Gerhard
, and
Anette
Rosenbach
.
2006
.
The winner takes it all—almost: Cumulativity in grammatical variation
.
Linguistics
44
:
937
971
.
Jardine,
Adam
.
2014
.
Computationally, tone is different
.
Ms., University of Delaware, Newark
.
Johnson,
C. Douglas
.
1972
.
Formal aspects of phonological descriptions
.
The Hague
:
Mouton
.
Jurgec,
Peter
.
2011
.
Icy targets
.
Ms., Meertens Institute, Amsterdam
.
Kager,
René
.
1999
.
Optimality Theory
.
Cambridge
:
Cambridge University Press
.
Kaplan,
Aaron F
.
2008
.
Noniterativity is an emergent property of grammar
.
Doctoral dissertation, University of California, Santa Cruz
.
Kaun,
Abigail
.
1993
. The coronal underspecification hypothesis. In
Papers in phonology
, ed. by
Daniel
Silverman
and
Robert
Kirchner
,
69
108
.
UCLA Occasional Papers in Linguistics 13
.
Los Angeles
:
UCLA, Department of Linguistics
.
Kaun,
Abigail
.
2004
. The phonetic foundations of the rounding harmony typology. In
Phonetically based phonology
, ed. by
Bruce
Hayes
,
Robert
Kirchner
, and
Donca
Steriade
,
87
116
.
Cambridge
:
Cambridge University Press
.
Kenstowicz,
Michael
.
1996
. Base-identity and uniform exponence: Alternatives to cyclicity. In
Current trends in phonology: Models and methods
, ed. by
Jacques
Durand
and
Bernard
Laks
,
363
393
.
Salford
:
University of Salford, ESRI
.
Rutgers Optimality Archive 103
. Available at http://roa.rutgers.edu.
Kenstowicz,
Michael
.
2009
.
Two notes on Kinande vowel harmony
.
Language Sciences
31
:
248
270
.
Kimper,
Wendell A
.
2011
.
Competing triggers: Transparency and opacity in vowel harmony
.
Doctoral dissertation, University of Massachusetts, Amherst
.
Kiparsky,
Paul
.
1981
.
Vowel harmony
.
Ms., MIT, Cambridge, MA
.
Kiparsky,
Paul
.
1985
.
Some consequences of Lexical Phonology
.
Phonology
2
:
85
138
.
Kiparsky,
Paul
.
2000
.
Opacity and cyclicity
.
The Linguistic Review
17
:
351
367
.
Kiparsky,
Paul
, and
Karl
Pajusalu
.
2003
.
Towards a typology of disharmony
.
The Linguistic Review
20
:
217
241
.
Krämer,
Martin
.
2003
.
Vowel harmony and correspondence theory
.
Berlin
:
Mouton de Gruyter
.
Krull,
Diana
,
Björn
Lindblom
,
Bang-er
Shia
, and
David
Fruchter
.
1995
. Cross-linguistic aspects of coarticulation: An acoustic and electropalatographic study of dental and retroflex consonants.
In Proceedings of the XIIIth International Congress of Phonetic Sciences
, ed. by
Kjell
Elenius
and
Peter
Branderud
,
3
:
436
439
.
Stockholm
:
KTH and Stockholm University
.
Ladefoged,
Peter
.
1964
.
A phonetic study of West African languages: An auditory-instrumental survey
.
Cambridge
:
Cambridge University Press
.
Langendoen,
D. Terence
.
1968
.
The London School of linguistics: A study in the linguistic theories of B. Malinowski and J. R. Firth
.
Cambridge, MA
:
MIT Press
.
Leben,
William
.
1973
.
Suprasegmental phonology
.
Doctoral dissertation, MIT, Cambridge, MA
.
Legendre,
Géraldine
,
Yoshiro
Miyata
, and
Paul
Smolensky
.
1990
.
Harmonic Grammar: A formal multilevel connectionist theory of linguistic well-formedness
.
Theoretical foundations. Technical Report, Department of Computer Science, University of Colorado, Boulder
.
Legendre,
Géraldine
,
Antonella
Sorace
, and
Paul
Smolensky
.
2006
. The Optimality Theory–Harmonic Grammar connection.
In The harmonic mind: From neural computation to Optimality-Theoretic grammar
, ed. by
Paul
Smolensky
and
Géraldine
Legendre
,
339
402
.
Cambridge, MA
:
MIT Press
.
Macdonell,
Arthur Anthony
.
1910
.
Vedic grammar
.
Strassburg
:
Trübner
.
Masica,
Colin P
.
1993
.
The Indo-Aryan languages
.
Cambridge
:
Cambridge University Press
.
McCarthy,
John J
.
1986
.
OCP effects: Gemination and antigemination
.
Linguistic Inquiry
17
:
207
263
.
McCarthy,
John J
.
2004
.
Headed spans and autosegmental spreading
.
Linguistics Department Faculty Publication Series 42
. Available at http://scholarworks.umass.edu/linguist_faculty_pubs/42.
McCarthy,
John J
.
2005
. Optimal paradigms. In
Paradigms in phonological theory
, ed. by
Laura
Downing
,
T. A.
Hall
, and
Renate
Raffelsiefen
,
170
210
.
Oxford
:
Oxford University Press
.
McCarthy,
John J
.
2008
.
The gradual path to cluster simplification
.
Phonology
25
:
271
319
.
McCarthy,
John J
.
2009a
.
Harmony in Harmonic Serialism
.
Linguistics Department Faculty Publication Series 41
. Available at http://scholarworks.umass.edu/linguist_faculty_pubs/41.
McCarthy,
John J
.
2009b
.
The P-map in Harmonic Serialism
.
Linguistics Department Faculty Publication Series 83
. Available at http://scholarworks.umass.edu/linguist_faculty_pubs/83.
McCarthy,
John J
.
2011
. Autosegmental spreading in Optimality Theory. In
Tones and features: Phonetic and phonological perspectives
, ed. by
John A.
Goldsmith
,
Elizabeth
Hume
, and
W. Leo
Wetzels
,
195
222
.
Berlin
:
Mouton de Gruyter
.
McCarthy,
John J.
, and
Alan
Prince
.
1993
.
Prosodic Morphology I: Constraint interaction and satisfaction
.
Ms., University of Massachusetts at Amherst and Rutgers University, Newark, NJ. Rutgers Optimality Archive 482
. Available at http://roa.rutgers.edu.
McPherson,
Laura
.
2016
.
Cumulativity and ganging in the tonology of Awa suffixes
.
Language
92
:
e38
e66
.
Monier-Williams,
Monier
.
1899
.
A Sanskrit-English dictionary
.
Oxford
:
Clarendon Press
.
Mullin,
Kevin
.
2011
.
Strength in harmony systems: Triggers and directional asymmetries
.
Ms., University of Massachusetts, Amherst. Rutgers Optimality Archive 1134
. Available at http://roa.rutgers.edu.
Myers,
Scott
.
1997
.
OCP effects in Optimality Theory
.
Natural Language and Linguistic Theory
15
:
847
892
.
Ní Chiosáin,
Máire
, and
Jaye
Padgett
.
1997
.
Markedness, segment realization, and locality in spreading
.
Technical Report LRC-97-01, Linguistics Research Center, University of California, Santa Cruz
.
Ní Chiosáin,
Máire
, and
Jaye
Padgett
.
2001
. Markedness, segment realization and locality in spreading. In
Segmental phonology in Optimality Theory
, ed. by
Linda
Lombardi
,
118
156
.
Cambridge
:
Cambridge University Press
.
Ohala,
John J.
, and
Manjari
Ohala
.
1993
. The phonetics of nasal phonology: Theorems and data. In
Nasals, nasalization, and the velum
, ed. by
Marie K.
Huffman
and
Rena A.
Krakow
,
225
249
.
San Diego, CA
:
Academic Press
.
Padgett,
Jaye
.
1995
. Feature classes. In
Papers in Optimality Theory
, ed. by
Jill
Beckman
,
Laura
Walsh Dickey
, and
Suzanne
Urbanczyk
,
385
419
.
University of Massachusetts Occasional Papers in Linguistics 18
.
Amherst
:
University of Massachusetts, Graduate Linguistic Student Association
.
Rutgers Optimality Archive 112
. Available at http://roa.rutgers.edu.
Pater,
Joe
.
2009a
.
Review of Smolensky and Legendre (2006), The harmonic mind
.
Phonology
26
:
217
226
.
Pater,
Joe
.
2009b
.
Weighted constraints in generative linguistics
.
Cognitive Science
33
:
999
1035
.
Pater,
Joe
.
2012
. Serial Harmonic Grammar and Berber syllabification. In
Prosody matters: Essays in honor of Elisabeth Selkirk
, ed. by
Toni
Borowsky
,
Shigeto
Kawahara
,
Takahito
Shinya
, and
Mariko
Sugahara
,
43
72
.
London
:
Equinox Press
.
Pater,
Joe
,
Rajesh
Bhatt
, and
Christopher
Potts
.
2007
.
Linguistic optimization
.
Rutgers Optimality Archive 924
. Available at http://roa.rutgers.edu.
Potts,
Christopher
,
Joe
Pater
,
Karen
Jesney
,
Rajesh
Bhatt
, and
Michael
Becker
.
2010
.
Harmonic Grammar with linear programming: From linear systems to linguistic typology
.
Phonology
27
:
77
117
.
Prince,
Alan
, and
Paul
Smolensky
.
1993
.
Optimality Theory: Constraint interaction in generative grammar
.
Technical Report, Rutgers University, Newark, NJ, and University of Colorado at Boulder
. Revised version,
Malden, MA
:
Blackwell (2004)
.
Renou,
Louis
.
1952
.
Grammaire de la langue védique
.
Paris
:
IAC
.
Renou,
Louis
.
1961
.
Grammaire sanscrite
.
Paris
:
Librairie d’Amérique et d’Orient Adrien-Maisonneuve
.
Rice,
Keren
, and
Peter
Avery
.
1991
. On the relationship between laterality and coronality. In
The special status of coronals
, ed. by
Carole
Paradis
and
Jean-François
Prunet
,
101
124
.
San Diego, CA
:
Academic Press
.
Ringen,
Catherine O
.
1976
. Vacuous application, iterative application, reapplication, and the unordered rule hypothesis. In
The application and ordering of grammatical rules
, ed. by
Andreas
Koutsoudas
,
55
75
.
The Hague
:
Mouton
.
Rose,
Sharon
, and
Rachel
Walker
.
2004
.
A typology of consonant agreement as correspondence
.
Language
80
:
475
531
.
Rose,
Sharon
, and
Rachel
Walker
.
2011
. Harmony systems. In
The handbook of phonological theory
, ed. by
John
Goldsmith
,
Jason
Riggle
, and
Alan C. L.
Yu
,
240
290
. 2nd ed.
Oxford
:
Blackwell
.
Sagey,
Elizabeth
.
1986
.
The representation of features and relations in non-linear phonology
.
Doctoral dissertation, MIT, Cambridge, MA
.
Schein,
Barry
, and
Donca
Steriade
.
1986
.
On geminates
.
Linguistic Inquiry
17
:
691
744
.
Selkirk,
Elisabeth O
.
1980
. Prosodic domains in phonology: Sanskrit revisited. In
Juncture
, ed. by
Mark
Aronoff
,
107
129
.
Saratoga, CA
:
Anma Libri
.
Shalev,
Michael
,
Peter
Ladefoged
, and
Peri
Bhaskararao
.
1993
.
Phonetics of Toda
. In
Fieldwork studies of targeted languages
,
89
123
.
UCLA Working Papers in Phonetics 84
. Available at http://escholarship.org/uc/item/8k45g432.
Shaw,
Patricia A
.
1991
. Consonant harmony systems. In
The special status of coronals
, ed. by
Carole
Paradis
and
Jean-François
Prunet
,
125
157
.
San Diego, CA
:
Academic Press
.
Simonsen,
Hanne G.
,
Inger
Moen
, and
Steve
Cowen
.
2000
.
Retroflex consonants in Norwegian: Are they really? Evidence from EMA and EPG
.
In Proceedings of the 5th Seminar on Speech Production
,
113
116
.
Smolensky,
Paul
.
1995
.
On the internal structure of the constraint component Con of UG
.
Handout from a talk at UCLA. Rutgers Optimality Archive 86
. Available at http://roa.rutgers.edu.
Smolensky,
Paul
, and
Géraldine
Legendre
.
2006
.
The harmonic mind: From neural computation to Optimality-Theoretic grammar
.
Cambridge, MA
:
MIT Press
.
Spajić,
Siniša
,
Peter
Ladefoged
, and
Peri
Bhaskararao
.
1996
.
The trills of Toda
.
Journal of the International Phonetic Association
26
:
1
21
.
Steriade,
Donca
.
1986
.
A note on coronal
.
Ms., MIT, Cambridge, MA
.
Steriade,
Donca
.
1995
.
Positional neutralization
.
Ms., UCLA, Los Angeles, CA
.
Steriade,
Donca
.
2000
. Paradigm uniformity and the phonetics-phonology boundary. In
Papers in laboratory phonology V: Acquisition and the lexicon
, ed. by
Michael B.
Broe
and
Janet B.
Pierrehumbert
,
313
334
.
Cambridge
:
Cambridge University Press
.
Steriade,
Donca
.
2009
. The phonology of perceptibility effects: The P-map and its consequences for constraint organization. In
The nature of the word: Essays in honor of Paul Kiparsky
, ed. by
Kristin
Hanson
and
Sharon
Inkelas
,
150
178
.
Cambridge, MA
:
MIT Press
.
Ussishkin,
Adam
.
1999
.
The inadequacy of the consonantal root: Modern Hebrew denominal verbs and output-output correspondence
.
Phonology
16
:
401
442
.
Vasu,
Srisa Chandra
.
1898
.
The Ashtádhyáyí of Páṇini, translated into English by Śríśa Chandra Vasu
,
vol. 8
.
Allahabad
:
Indian Press
.
Vergnaud,
Jean-Roger
, and
Morris
Halle
.
1978
.
Metrical structures in phonology
.
Ms., MIT, Cambridge, MA
.
Wackernagel,
Jacob
.
1896
.
Altindische Grammatik
,
vol. 1
.
Göttingen
:
Vandenhoeck & Ruprecht
.
Wackernagel,
Jacob
, and
Albert
Debrunner
.
1957
.
Altindische Grammatik
.
Gottingen
:
Vandenhoeck & Ruprecht
.
Walker,
Rachel
.
2005
.
Weak triggers in vowel harmony
.
Natural Language and Linguistic Theory
23
:
917
989
.
Walker,
Rachel
.
2014
.
Nonlocal trigger-target relations
.
Linguistic Inquiry
45
:
501
523
.
Walker,
Rachel
,
Dani
Byrd
, and
Fidèle
Mpiranya
.
2008
.
An articulatory view of Kinyarwanda coronal harmony
.
Phonology
25
:
499
535
.
Walker,
Rachel
, and
Fidèle
Mpiranya
.
2005
. On triggers and opacity in coronal harmony. In
BLS 31: General Session and Parasession on Prosodic Variation and Change
, ed. by
Rebecca T.
Cover
and
Yuni
Kim
,
383
394
.
Berkeley
:
University of California, Berkeley Linguistics Society
. Available at http://journals.linguisticsociety.org/proceedings/index.php/BLS/issue/view/30.
Whitney,
William Dwight
.
1889
.
Sanskrit grammar, including both the Classical language and the older dialects, of Veda and Brahmana
. 2nd ed.
Cambridge, MA
:
Harvard University Press
.
Wilson,
Colin
.
2003
.
Analyzing unbounded spreading with constraints: Marks, targets, and derivations
.
Ms., UCLA, Los Angeles, CA
.
Yip,
Moira
.
2002
.
Tone
.
Cambridge
:
Cambridge University Press
.
Zuraw,
Kie
.
2007
.
The role of phonetic knowledge in phonological patterning: Corpus and survey evidence from Tagalog
.
Language
83
:
277
316
.
Zuraw,
Kie
.
2013
.
*MAP constraints
.
Ms., UCLA, Los Angeles, CA
.