This paper focuses on the Japhug language (local name /kɯrɯ skɤt/) of Kamnyu village (/kɤmɲɯ/, Chinese Ganmuniao 干木鸟) in Gdongbrgyad area (/ʁdɯrɟɤt/, Chinese Longerjia 龙尔甲), Mbarkhams county (Chinese Maerkang 马尔康), Rngaba prefecture, Sichuan province, China. Japhug belongs to the Sino-Tibetan family, and is one of the four Rgyalrong languages, alongside Tshobdun, Zbu and Situ.Footnote 1
The description is based on the author’s fieldwork. The word lists and the short story in the ‘Transcription of a recorded passage’ section have been provided by Tshendzin (Chenzhen 陈珍, female, born in 1950), a retired schoolteacher (a native speaker of Japhug, bilingual in Sichuan Mandarin since childhood).
Japhug has a highly developed system of ideophones (Jacques Reference Jacques2013b), which present unusual phonological features, in particular rare clusters. In the following discussion, phonemes or clusters found exclusively on ideophones will be treated separately. In addition, about a quarter of the Japhug vocabulary is borrowed from Tibetan, and these loanwords, like the ideophones, fill some gaps in the phonotactic distribution of vowels and consonants (on gap-filling by loanwords see Martinet Reference Jacques1955: 63–64). These cases are carefully distinguished from the native vocabulary in the analyses that follow, in order to bring out the phonotactics of inherited Japhug vocabulary.
Consonants
In Japhug, syllables follow the template (C)(C)(C)V(C) or (C)(C)(C)V(V) with initial clusters containing at most three consonants, and at most one coda. Given the complexity of possible onsets, it is not practical, in the case of Japhug, to provide an exhaustive list of possible syllables in the language (unlike Naish languages for instance, see Michailovsky & Michaud Reference Jacques2006, Michaud Reference Lai2012).
Simple onsets
The consonant inventory of Japhug comprises 50 phonemes. There is a general four-way contrast in stops and affricates between unvoiced unaspirated, unvoiced aspirated, voiced and prenasalized.
Since monosyllabic words are few, most of the example words provided here involve disyllabic words, whose first syllable illustrates the consonant at issue followed by the vowel /ɯ/, the most common one. For some examples involving possessed nouns (on this topic see Jacques Reference Jacques2014:4), a possessive prefix must always be present, and we always give here the third person singular prefix /ɯ-/.
Among the consonants of Japhug, four are only attested in borrowings from Tibetan and/or ideophones: /ʂ/, /dʐ/, /dʑ/ and /g/.
The analysis of prenasalized voiced stops and affricates, palatal stops and /ɬ/ as unitary phonemes rather than clusters /NC/, /c+j/ and /l+x/, respectively, will be justified below.
The /ɬ/ is slightly aspirated [ɬʰ], unlike other unvoiced fricatives (note that many languages in the area have constrastive aspirated fricatives, see Jacques Reference Jacques2011; Japhug however has no such contrast).
As in many languages of the Tibetan area, the /r/ is a trilled retroflex voiced fricative [
] in onset position, sometimes realized as a simple voiced fricative [ʐ].
Consonants clusters
Japhug boasts 415 clusters in syllable onset position: 315 clusters with two consonants and 100 with three consonants. Clusters that are only possible at syllable boundaries are not included in this count.
Japhug being a heavily prefixing language (on which see Jacques Reference Jacques2013a), most of its noun or verb stems are prefixed, and thus a considerable part of onset clusters are not attested word-initially. For instance, the cluster /zmb/ is only found in the word /tɤzmbɯr/ ‘silt’, which contains a nominal prefix /tɤ-/ (see Table 1 below).
Table 1 List of consonant clusters with /w/ as a first element (15+8).
Yet, speakers are able to parse words into syllables; in the case of /tɤzmbɯr/ ‘silt’, the only possible syllabification in /tɤ|zmbɯr/, not */tɤz|mbɯr/, and thus we can ascertain that /zmb/ can be counted as a possible onset in Japhug. On the other hand, in examples like /pjɤnɯndzɯlŋɯz/ ‘he dozed off’, both syllabifications /pjɤ|nɯ|ndzɯl|ŋɯz/ and /pjɤ|nɯ|ndzɯ|lŋɯz/ are possible, so that /lŋ/ is not counted among syllable onset clusters in Japhug.
Decisive evidence from partial reduplication
A useful test to analyze and classify clusters is partial reduplication, a very productive process which can be applied to both verb and noun stems, and has a variety of morphosyntactic functions (see Jacques Reference Jacques2007). When partial reduplication is applied to a syllable, the rhyme of the replicated syllable is changed to /ɯ/ in the replicant.
Some clusters are affected by the partial reduplication: when the last consonant of a cluster is one of the non-nasal sonorants (/r/, /l/, /j/, /w/, /ɣ/ or /ʁ/), and the preceding consonant in neither a sonorant nor a sibilant fricative, the sonorant is deleted, as in the derivation /ɲɤ-prɤt/ ‘he cut it’ /ɲɤ-nɤ-pɯ~prɤt/ ‘he cut it in all directions’.Footnote 2
When the penultimate consonant of the cluster is a sonorant and the last consonant is a non-nasal sonorant which is not a glide (/r/, /l/, /ɣ/ or /ʁ/), this last consonant is not deleted, as in /ko-wraʁ/ ‘he attached it’ /ko-nɤ-wrɯ~wraʁ/ ‘he attached it in all directions’.
When the prenultimate consonant is a sibilant fricative (/s/, /z/, /ɕ/, /ʑ/) and the last consonant is a non-nasal sonorant which is not a glide (/r/, /l/, /ɣ/ or /ʁ/), there are various possibilities, which are detailed in below.
This morphophonological rule is thus crucial in analyzing and classifying consonant clusters. Sonorants that undergo deletion when partial reduplication is applied are henceforth designated as medial consonants, and it is postulated that they do not belong to the same constituent as the rest of the onset.
In the remainder of this section, we present a complete inventory of consonant clusters in Japhug. In Tables 1–12, groups only attested in Tibetan loanwords or ideophones (or deideophonic verbs), and not in the native vocabulary, are indicated in light gray and gray, respectively. We only count clusters in syllable onsets, not clusters occurring only between syllable boundaries, some of which will be treated below.
Clusters not ending in a (non-nasal) sonorant
Clusters whose last consonant is not a non-nasal sonorant have a limited number of possible consonants in first position: /w/, /s/, /z/, /ɕ/, /ʑ/, /l/, /r/, /ʂ/, /j/, /ɣ/, /x/, /ʁ/, /χ/, /n/, /m/ and the homorganic nasal, except for a few clusters in stop+ /ɕ/.
Clusters beginning in /w/ are listed in Table 1. /w/ is realized as [f] or [ɸ] before unvoiced obstruents and as [v] or [β] before voiced ones. /w/ does not appear before nasal or prenasalized segments, and cannot be followed by another labial consonant. Some clusters with /w/+voiced obstruents (/wz/ and /wg/) are only attested in Tibetan loanwords. Clusters with three consonants whose first element is /w/ and the last one is not a sonorant are all restricted to Tibetan borrowings except /wxt/, which is realized as [xʷt], with a labio-velarized fricative (and labializes preceding /ɯ/ and /ɤ/ to [u] and [o], respectively). Not all speakers maintain the contrast between /wxt/ and /xt/, and the former cluster is only attested in a single word /wxti/ ‘be big’.
Clusters beginning in an alveolar fricative /s/ and /z/ are listed in Table 2. /s/ and /z/ as the first element of a cluster are only contrastive before a sonorant. With obstruents, the fricative has the same voicing value as the following consonant. All clusters of this type are attested in the native vocabulary.
Table 2 List of consonant clusters with /s/ or /z/ as a first element (23).
Clusters with /l/ as the first element are listed in Table 3. /l/ is attested before all places of articulation, but not before coronal fricatives.
Table 3 List of consonant clusters with /l/ as a first element (17+1).
Clusters with /r/ and /ʂ/ as the first element are listed in Table 4. /r/ and /ʂ/ are almost in complementary distribution as the first element of a cluster, the former appearing before voiced consonants and the latter after unvoiced ones (except before /ɣ/). In keeping with this generalization, /r/+nasal clusters are widely attested (nasals are phonemically and phonetically voiced in Japhug), while /ʂ/+nasal clusters are only attested in some ideophones.
Table 4 List of consonant clusters with /r/ or /ʂ/ as a first element (35).
There are some phonotactic constraints on the distribution of these consonants: and /r/ and /ʂ/ never appear before retroflex fricatives and affricates.
The glide /j/ and the alveolo-palatal fricatives /ɕ/ and /ʑ/ (Table 5) only occur before labial, dental, velar and uvular stops; they are marginally attested with retroflex affricates. The /j/ glide, unlike other consonants, neither devoices nor fricativizes when occurring as the first element of a cluster whose second element is an obstruent.
Table 5 List of consonant clusters with /j/, /ɕ/ or /ʑ/ as a first element (29+1).
Clusters with /x/ and /ɣ/ as the first element are listed in Table 6. The velar fricatives always share their voicing feature with the following segment. They are compatible with all places of articulation except velars and uvulars within a single onset. Clusters combining uvular and velar consonants are, however, possible in heterosyllabic clusters.
Table 6 List of consonant clusters with /x/ or /ɣ/ as a first element (23).
Clusters with /χ/ and /ʁ/ as the first element are listed in Table 7. The uvular fricatives always share their voicing feature with the following segment when it is an obstruent. With nasal sonorants, they are almost always voiced except in the group /χɲ/, which contrasts with /ʁɲ/ and is only attested in ideophones.
Table 7 List of consonant clusters with /χ/ or /ʁ/ as a first element (26).
Clusters with nasal segments as the first element (not counting voiced prenasalized stops and affricates) are listed in Tables 8 and 9. We find homorganic nasal clusters, compatible with all places of articulation, and non-homorganic ones, which can be either clusters in /n/+labial or velar or /m/+non-labial. There are no clusters with a nasal directly followed by a fricative or any non-nasal sonorant. After /n/ and /m/, the contrast between voiced preinitial and voiced consonants is neutralized. Since only the prenasalized phoneme /ɴɢ/ exists while no simple voiced /ɢ/ is attested, the cluster [mɢ] can only be analyzed as /mɴɢ/ phonologically.
Table 8 List of consonant clusters with a homorganic nasal or /m/ as first element (38+1).
Table 9 List of consonant clusters with non-homorganic /n/ as first element (4).
Finally, we find a few clusters comprising a stop followed by the fricative /ɕ/: /pɕ/ (as in /ɯ-pɕi/ ‘outside’), /kɕ/ (as in /kɯjkɤkɕi/ ‘marten’), /lpɕ/ and /mpɕ/ (examples of the latter two are found in Tables 3 and 8). These clusters historically originate from aspirated stops followed by /j/ (/pʰj/, /kʰj/), but their pattern in reduplication indicates that this is no longer true synchronically. For instance, /mpɕɤr/ ‘it is beautiful’ is reduplicated as /mpɕɯ~mpɕɤr/) not /mpʰɯ~mpʰʲɤr/.
Clusters ending in a (non-nasal) sonorant
The non-nasal sonorants (/r/, /l/, /j/, /w/, /ɣ/ or /ʁ/) can occur after any consonant except nasals. In this section, clusters are listed by the penultimate consonant (the one immediately preceding the non-nasal sonorant).
The list of all clusters whose final consonant is a glide /j/ or /w/ is provided in Table 10. The glides /j/ or /w/ are medials in all clusters except /wj/, /jw/, /ɣj/ and /ʁj/. The labio-velar /w/ has a very restricted distribution as last element of a cluster; in the native non-ideophonic vocabulary, it only occurs after /l/, /z/ and /j/ and is never found in clusters comprising three consonants. The palatal glide /j/ has a wider distribution: it occurs after all places of articulation except palatal and retroflex.
Table 10 List of consonant clusters ending in /j/ or /w/ (30+18).
Table 11 provides a list of all clusters whose final consonant is a liquid /r/ or /l/. Clusters ending in /r/ cannot contain another /r/ or /l/ segment, or any retroflex consonant (on the crosslinguistic rarity of the cluster /lr/, see Baroni Reference Baroni2014: 78). Clusters ending in /l/ never contain another /l/, but allow the presence of /r/ (/rl/, /rɴɢl/). The sonorants /r/ and /l/ are medials before all stops and affricates, as well as before the voiced fricatives /z/ and /ʑ/. In all other contexts, they are not medials.
Table 11 List of consonant clusters ending in /l/ or /r/ (44+52).
Table 12 provides a list of all clusters whose final consonant is a dorsal voiced fricative /ɣ/ or /ʁ/. Only one dorsal segment is possible within an onset-initial consonant cluster. The sonorants /ɣ/ and /ʁ/ are medials only before all stops and affricates.
Table 12 List of consonant clusters ending in /ɣ/ or /ʁ/ (32+18).
The counts of all initial consonant clusters in Japhug are summarized in Table 13.
Table 13 Count of consonant clusters.
Heterosyllabic clusters
The list of clusters presented in the previous section only include syllable-initial clusters; many more additional clusters are possible across syllable (and morpheme) boundary. Clusters made of the coda of the first syllable and the onset of the second syllable of a disyllabic word that are not attested in simple onsets can be grouped in three categories.
First, while in syllable onsets we never find clusters containing two stops, such clusters are found across syllable boundary. In Tibetan loanwords, two clusters with /t/ as the first element, /tk/ and /tp/, are attested in words such as /χɕit.ka/ ‘spring’ (Tibetan /dpʲid.ka/) and /rɟɤt.pa/ ‘eighth month’ (Tibetan /brgʲad.pa/).
The only other cluster containing two stops, /pt/, is attested in the word /sqap.tɯɣ/ ‘eleven’. This is the only case of /p/ appearing as coda (instead of /w/) in a non-ideophonic word.
Second, although Tables 6 and 7 above show restricted possibilities for clusters whose first element is a velar or a uvular fricative, in heterosyllabic clusters there are few constraints on the phonotactics of these clusters. A velar fricative can be followed by a uvular fricative: the cluster /ɣʁ/ is attested in some speakers between morpheme boundaries in the case of the causative of intransitive verbs with the onset /ʁ/. For instance, the verb /ʁaʁ/ ‘hatch’ has a causative /sɯ-ʁaʁ/ or /sɯɣ-ʁaʁ/ depending on the speaker. Uvular fricatives followed by velar fricatives or stops are also attested, as in /praʁ.kʰaŋ/ ‘cave’ for example.
Third, while in homosyllabic clusters nasal elements cannot be followed by fricatives or non-nasal sonorants, as shown in Table 8 above, such combinations are possible in heterosyllabic clusters, such as in /naŋ.ʁɯ/ ‘shirt’, /to.nɤ.tsɯm.ɣɯt/ ‘he brings it here and there’, /ɲɯ.nɯ.sɯm.ʁɲɯz/ ‘he hesitates’, /kʰoŋ.rɤl/ ‘hollow tree’.
The sonority sequencing principle in Japhug
Many works in phonology support the idea that all segments of the world’s languages follow a universal scale of sonority (for instance Vennemann 1988, Blevins 1995, Parker 2002, Baroni 2014; see Ohala 1990 for an opposing view). A particularly elaborate version of the sonority hierarchy has been proposed by Parker (Reference Parker2002: 235):
(1) low vowels > mid vowels > high vowels > /ə/ > glides > laterals > flaps > trills > nasals > /h/ > voiced fricatives > voiced stops > voiceless fricatives > voiceless stops and affricates
The notion of sonority is used in particular to account for observed generalizations in the structure of consonant clusters: in most languages, clusters follow the so-called sonority sequencing principle (SSP, Blevins Reference Blevins and Goldsmith1995: 200):
(2) Between any member of a syllable and the syllable peak, a sonority rise or plateau must occur.
According to this hierarchy, in onset clusters, sonorants are expected to be closer to the syllable nucleus than obstruents (/prV/ is favoured over /rpV/), and glides are expected to be closer to the nucleus than any other consonant (/ljV/ is favoured over /jlV/).
Onset clusters with glides or non-nasal sonorants violating the SSP (/jC/, /wC/, /lC/ and /rC/, where C is a consonant lower in the sonority hierarchy) are crosslinguistically rare, but have been documented for instance in Oceanic languages such as Dorig and Hiw (Francois Reference Francois2010: 405-412).
Japhug, like other Gyalrongic languages,Footnote 3 is rich in SSP-infringing clusters. For instance, no fewer than seven clusters with a labial consonant and a non-nasal sonorant violating the SSP are attested: /lp/, /jp/, /lm/, /rm/, /jm/, /rmb/ and /jmŋ/.
More surprisingly, we find clusters violating the SSP without equivalent non-violating clusters. For instance, out of 15 di-consonantal clusters with /j/ as the first element (all SSP-infringing except for /jw/), eight have no equivalent SSP-compliant cluster (/jm/, /jt/, /jn/, /jtsʰ/, /jtʂʰ/, /jndʐ/, /jŋ/, /jχ/).
Clusters vs. unitary phonemes
In this section, we justify the need to analyze two groups of consonants as unitary phonemes rather than clusters, namely the prenasalized voiced stops, the palatals and /ɬ/.
The prenasalized voiced stops and affricates /mb/, /nd/, /ndz/, /ndʑ/, /ndʐ/, /ɲɟ/, /ŋg/ and /ɴɢ/ have unvoiced and unvoiced aspirated counterparts such as /mp/, /nt/, /nts/, /ntɕ/, /ntʂ/, /ɲc/, /ŋk/ and /ɴq/. Yet, there are two pieces of evidence showing that the prenasalized voiced stops and affricates differ from the prenasalized unvoiced ones.
First, the former can appear in clusters preceded by fricatives or non-nasal sonorants, as in /ʑmbr/, /jndʐ/ or /rɴɢl/, while the latter cannot. Clusters such as * /ʑmpr/, * /jntʂ/ or * /rɴql/ are not tolerated in Japhug.
Second, the uvular voiced prenasalized /ɴɢ/ has no simple voiced counterpart * /ɢ/, which therefore precludes analyzing /ɴɢ/ as a cluster /n+ɢ/.
The palatal stops /c/, /cʰ/, /ɟ/ and /ɲɟ/ in Japhug cannot be analyzed as velar+ /j/ clusters, as a clear contrast exists between the palatal series and a velar followed by /j/, in minimal pairs such as /pɯ-ɲɟo/ ‘he had damages’ and /pɯ-ŋgjo/ ‘he slipped’. The differing syllabic structure of the onsets /ɲɟ/ and /ŋgj/ is confirmed by their reduplication patterns: while in the former the palatalization is present on the replicant /pɯ-nɤ-ɲɟɯ~ɲɟo/ ‘he had damages everywhere’, in the latter the /j/ is not replicated (/pɯ-nɤ-ŋgɯ~ŋgjo/ ‘he slipped everywhere’).
Japhug presents an impressively high number of palatalization contrasts among coronal and dorsal onsets. All contrasts are illustrated in Table 14, with the vowel /o/.
Table 14 Palatalization contrasts among coronal and dorsals (the relevant consonants are set in bold).
The unvoiced lateral /ɬ/ is a marginal phoneme in Japhug, which does not appears in clusters (except heterosyllabic ones, as in /cɯɣɬaj/ ‘symptom whereby the oral cavity becomes white’) and is very rare in the native vocabulary. Yet, its phonemic status is justified by the fact that it contrasts with /lx/; there are no minimal pairs between the two, but the contrast can be indirectly illustrated by examples such as /alxaj/ ‘(his clothes) are not properly put on’ /lxɯlxi/ ‘thick and cumbersome’ on the one hand, and /pjɤɬɤt/ ‘he became old’ and /ɬɤndʐi/ ‘ghost’ on the other hand.
Codas
As is common in the Sino-Tibetan family, the inventory of consonants in coda position in Japhug is more restricted than in initial position. Only twelve consonants appear in coda position: /-p/, /-w/, /-m/, /-t/, /-z/, /-n/, /-l/, /-r/, /-j/, /-ɣ/, /-ŋ/ and /-ʁ/. The stop /-p/ is only restricted to a few ideophones, and is not found in the inherited non-ideophonic vocabulary and in Tibetan loanwords (except as the first element of the heterosyllabic cluster /pt/ in the word /sqap.tɯɣ/ ‘eleven’). The codas /-n/, /-l/ and /-ŋ/ are extremely rare (but not entirely absent) in the non-ideophonic native vocabulary.
In word-final position, codas are voiced when followed by a word beginning with a voiced consonant or a vowel, but are devoiced in phrase-final position, before a pause or before a voiceless segment.
In isolation, word-final /-z/, /-r/, /-j/, /-ɣ/ and /-ʁ/ in particular are realized as [s], [
], [
], [x] and [χ], respectively, as in the examples in Table 15. The coda /-ʁ/ can also be realized alternatively as pharyngealization of the preceding vowel.
Table 15 Realization of non-nasal codas in Japhug.
Vowels
Japhug has eight vowel phonemes.
The mid-open unrounded vowels /ɤ/ and /e/ are only marginally contrastive: /ɤ/ does not occur in word-final open syllables except in unaccented clitics (like the linker /nɤ/ ‘if’), and /e/ only occurs in the last (accented) syllable of a word. They are only contrastive with the coda /-t/.
The vowel /y/ is only found with some speakers in the word ‘fish’ and the verbs derived from it. It is nevertheless contrastive with /ɯ/ and /u/ (as shown by the quasi-minimal pairs /qaɟy/ ‘fish’, /waɟɯ/ ‘earthquake’ and /ɟuli/ ‘flute’). Other speakers pronounce ‘fish’ with a medial /w/ as /qaɟwi/.
When followed by a syllable containing a rounded vowel (/u/ or /o/), the back unrounded vowels /ɯ/ and /ɤ/ optionally undergo rounding harmony to [u] and [o], respectively (for instance /ɣdɤso/ ‘grub’ in Table 6 above can be pronounced as [ɣdoso]).
Rhymes
There are strong phonotactic constraints on possible rhymes in Japhug. The only coda attested with all vowels is /-t/ (see Tables 16 and 17); the rhymes /-et/, /-yt/ are only attested in the perfective second singular forms, which have a suffix /-t/ in the variety of Japhug under study.
Table 16 List of possible rhymes in Japhug.
Table 17 Examples of closed syllable rhymes in /-t/.
Before /-j/, the contrasts between /ɯ/ and /i/ on the one hand, and /ɤ/ and /e/ on the other hand, are neutralized. The rhyme /-aj/ is realized as [ɛj].
In closed syllables with an alveolo-palatal or a palatal consonant preceding the vowel, the vowel /ɯ/ is fronted and its contrast with /i/ is neutralized in nearly all positions. It is only maintained before /-t/ in forms with the past transitive suffix /-t/. For instance, we find the minimal pair /tɤ-tɯ-cɯ-t/ ‘you opened it’ (pfv-2-open-pst) and /lɤ-tɯ-cit/ ‘you moved’ (pfv-2-move).
Suprasegmentals
Unlike other Rgyalrong languages (Sun Reference Sun2005, Lin Reference Lin2012), Japhug has no tonal contrasts. However, there is morphologically determined stress. Phonological words only have one stress, which is located by default on the final syllable of the word (regardless of its part of speech).
The personal agreement suffixes and the evidential suffix /-ci/ never receive stress, and their vowels are optionally devoiced. For instance, /tɤ-ndza-t-a/ ‘I ate it’ (pfv-eat-pst-1sg) is realized as [tɤndzát
] or [tɤndzáta]. In verbal forms with these prefixes, stress is penultimate, or even antepenultimate in the case of verb forms with two suffixes, as in /to-k-ɤmɯ-rpú-ndʑɯ-ci/ ‘they bumped into each other’ (ifr-evd-recip-bump-du-evd).
Only two verbal prefixes, the inverse /-wɣ-/ and the negative testimonial /m
j-/, attract stress, as in /pɯ-t-wɣ-mto/ ‘he saw you’ and /mj-ndze/ ‘he does not eat it’.
The only other morphological process in the language that influences stress is the comitative adverbs formation. These adverbs meaning, ‘together with X’, are built by combining the stress-bearing /k
-/ prefix with a reduplicated noun stem, as in the examples in Table 18.
Table 18 Examples of associative adverbs in Japhug.
Transcription of a recorded passage
This text was translated into Japhug from the Chinese version of Aesop’s fables by Chenzhen 陈珍 /tsʰɯndzɯn/ in May 2014. In the passage below, each segment includes lines with phonemic transcription, English word-for-word gloss, and English idiomatic translation. Glosses follow the Leipzig Glossing Rules, except for the following: auto = autobenefactive/spontaneous, emph = emphatic, genr = generic, ifr = inferential, inv = inverse, lnk = linker, testim = testimonial.
Acknowledgements
I wish to thank Alexis Michaud and two anonymous reviewers for comments on previous versions of this work; I am responsible for any remaining errors. This research was funded by the HimalCo project(ANR-12-CORP-0006) and is related to the Labex EFL (PPC2 Evolutionary approaches to phonology: New goals and new methods (in diachrony and panchrony)).
