====== Asmat-Muli Strait ====== Timothy Usher, Santa Fe Institute ===== Situation ===== The Asmat-Muli Strait family consists of perhaps 11 languages spoken along New Guinea's southwest coast from Kamrau Bay in the northwest to Kolopom and Komolom islands to the southeast. ===== Subclassification ===== The internal classification of Asmat-Muli Strait languages is as follows:\\ \\ Asmat-Muli Strait Muli Strait Konorau Mombum Asmat-Kamrau Bay Kamrau Bay Asmat-Kamoro Kamoro Sempan Asmat\\ \\ The differences between Muli Strait and Asmat-Kamrau Bay are significantly greater than those between Kamrau Bay and Asmat-Kamoro. The Muli Strait languages, Konorau and Mombum, are very similar to one another (q.v. Geurtjens 1933: 398-433.). ===== Sources ===== **[under construction]**\\ \\ Geurtjens (1933: 398-433) 455 comparative terms for Komelom(sch) (Mombum) and Koneraw(sch) (Konorau) Drabbe (1950: 561-574) brief grammar and 422 comparative terms for Mombum Voorhoeve (1980: 61-121) Proto-Asmat includes comparisons to Drabbe's Mombum\\ \\ Sources for Asmat-Kamrau Bay languages are listed in their respective pages. ===== History of classification ===== The special relationship between Drabbe's (1953) Central South Coast (Asmat-Kamoro) and Mombum was first recognized by Voorhoeve (1968: 3-4,) who included Mombum in the Asmat-Awyu-Ok Family of his Central and South New Guinea Stock: "Mombum is a border case; its closest relationship is with Asmat. Since it does not show comparably close links with any other language it has been provisionally included in the Asmat-Awyu-Ok Family." This assertion was reiterated in Voorhoeve's (1980) Asmat survey and reconstruction, where he states (p. 1), "The Mombum Family is the closest relative of the Asmat-Kamoro Family; together they form a subgroup with the CSNG [Central and South New Guinea] Stock," and a number of Mombum comparisons are presented along with Proto Asmat (pp. 61-121.)\\ \\ While Bromley (1973: 18) credits Voorhoeve’s proposal, it was inexplicably ignored in works which otherwise recapitulated Voorhoeve’s findings, such as Wurm (1982) and other tertiary works which followed. Perhaps Voorhoeve’s additional insistence upon a special relationship between Asmat, Awyu-Dumut and Ok had confused matters. Both Asmat-Muli Strait and Digul River-Ok form valid subgroups of Trans New Guinea, but whether these two in turn form a higher level subgroup is a much hazier proposition. The sharp division between Asmat-Kamoro on the one hand and Awyu-Dumut and Ok on the other is quite visible in Voorhoeve (2005: 152-164,) although his conclusion (pp. 164-165) is somewhat equivocal.\\ \\ Wurm (1982: 136-138) presents the Mombum Family as a first order member of the Central and South New Guinea Stock along with Asmat-Kamoro (including Kamrau Bay) and eight other families, but notes typological similarities between the two groups (p. 139): "The second type is represented by the languages of the Mombum and the Asmat-Kamoro Families. It is characterized by the presence of two genders in third (in some instances also second) person pronouns, the appearance of subject and object suffixes with the verb, and a low-level elaboration of sentence-medial verb forms in which identity versus non-identity of the two subjects involved is usually not distinguished by separate forms." It's not immediately clear to what evidence Wurm refers in his comments about pronominal genders, but the formal as well as typological similarities between verbal desinences (below) constitute strong evidence which went unmentioned in Voorhoeve (1968, 1980.)\\ \\ Greenberg (1971: 840) recognized the relationship between Kamrau Bay and Asmat-Kamoro, but the Muli Strait languages were placed with a number of unrelated nearby languages in his Jei subgroup of South New Guinea (pp. 829-830,) probably due to loans shared with Kimaghama immediately to the north (q.v. Geurtjens 1933: 398-433.) Asmat-Kamrau Bay is left unplaced relative to his larger Indo-Pacific scheme (pp. ….) ===== Historical phonology ===== **[under construction]**\\ \\ Proto-Asmat-Muli Strait had perhaps 12 consonants and 6 vowels as follows: |< - 60px 60px 60px 60px >| | *m | *n | | | | *p | *t | *s | *k | | *b | *d | | *g | | *w | *ɾ | *j | | |< - 60px 60px 60px >| | *i | | *u | | *e̝ | | *o̝ | | | | *ɔ | | | *a | | …\\ \\ …\\ \\ Initial consonants correspond as follows: |< - 80px 80px 80px >| | Asmat-Muli | A,-Kamrau | Muli Strait | | *m- | *m | *m | | *n- | *n | *n | | *p- | *ɸ | *… | | *t- | *s | *t | | *s- | *s | *s | | *k- | *ø | *k | | *b- | *b | *p | | *d- | *ɟ | *t | | *g-/i e | *g | *j | | *g- | *ø | *ɣ | | *w- | *w | *p | | *j- | *j | *z | Medial consonants correspond as follows: |< - 80px 80px 80px >| | Asmat-Muli | A,-Kamrau | Muli Strait | | *-m- | *m | *m | | *-n- | *n | *n | | *-p- | *… | *… | | *-t- | *s | *t | | *-s- | *s | ? | | *-k- | *k | ? | | *-b- | *b | ? | | *-d- | *ɾ | *d | | *-g-/i e | *g | *ø | | *-g- | *ø | *ɣ | | *-w- | *w | *b | | *-ɾ- | *ɾ | *r | | *-j- | *j | *r | Subfinal consonants correspond as follows: |< - 80px 80px 80px >| | Asmat-Muli | A,-Kamrau | Muli Strait | | *-mV | *mV | *m | | *-nV | *nV | *ɾ | | *-pV | *pV | *p | | *-ɸV | *ɸV | *p | | *-tV | *tV | *ɾ | | *-sV | *sV | *ɾ | | *-kV | *… | *… | | *-bV | *bV | | | *-dV | *… | *… | | *-gV/i e | *gV | *ɣ | | *-gV | *ø | *ɣ | | *-wV | *wV | *b | | *-ɾV | *-ɾV | *ø | | *-jV | *… | *… | \\ …:\\ \\ …:\\ \\ These correspondences are exemplified as follows.\\ \\ \\ \\ Probably any consonant but apical non-stop /*r/ can occur initially, although several probable initial correspondences have not yet been exemplified for both families.:\\ \\ Initial bilabial nasal /*m/ is retained as such in both families: |< - 100px 100px 100px 100px >| | |Asmat-Muli |Asmat-Kamrau|Muli Strait| | |***m-** |***m** |***m** | |saliva/spittle|*mVtapu |*masa[p/ɸ]u |*mitap | |thought |*mVnipi |*minipi |*me̝nip | |nose/tip |*mVnVgV |*m[e/a]n[e] |*mæne̝ɣ | |knowledge |*mat[ɔ]m[ɔ]|*mas[a/o]mo | | |foot |*mawu |*mawu |*mɔb | |water |*mVi |*moi |*mo̝i | Iinitial apical nasal /*n/ is retained as such in both families: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***n-** |***n** |***n** | |eat |*nV |*nV |*no̝ku | |1 sg.|*nV |*no |*no̝ | Initial bilabial voiceless stop /*p/ …: |< - 100px 100px 100px 100px >| ||Asmat-Muli|Asmat-Kamrau|Muli Strait| ||***p-** |***ɸ** |**** | || | | | Initial apical voiceless stop /*t/ …: |< - 100px 100px 100px 100px >| ||Asmat-Muli|Asmat-Kamrau|Muli Strait| ||***t-** |***s** |***t** | || | | | Initial laminal voiceless /*s/ is retained as such in both families: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***s-** |***s** |***s** | |sand |*sini |*sini |*sir | |tooth/sharp |*sisV |*sisV |*-sir | |charcoal/black|*sosV |*sosV |*sɔr | The third of these deserves some comment: Voorhoeve (1980: 68-69, 74) compres Mombum [sor] "black," which aslo means "charcoal (q.v. Geurtjens 1933: 418-419, 420-421) to both Asmat /*sosake/ "black" and Asmat /*sose/, which Voorhoeve gives as "blackening, scorching, torch of dry palm leaves," presumably under the assumption that the former is derived from the latter through suffixation. However, Drabbe (1953: 99, 1963: 220-221) translates this root as "fakkel" ("torch",) and a look through the Cental Asmat dictionaries available to us supports only the meaning "torch of dry palm leaves" (Drabbe 1958: 102, Voorhoeve 1965: 350, 352, Roesler and Roesler 2011: 154.) Thus, while it appears to correspond to the Muli strait word, the semantics are more problematic than they appear to be in Voorhoeve's presentation.\\ \\ Initial velar voiceless stop /*k/ …: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***k-** |***ø** |***k** | |do/make |*ke-mV- |*e-mV- |*ke̝mV- | |sago pulp|*kapimi |*apimi | | |crocodile|*kewV |*ewo | | |unripe |*kɔFɔ |*oFo |*kɔp | Initial bilabial voiced stop /*b/ is devoiced to /*p/ in Muli Strait as it is in Asmat-Kamoro: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***b-** |***b** |***p** | |see/look |*pVgVɾV- |*boɾa- |*pe̝ɣe̝r | |flatulate|*bo̝ |*bu- |*po̝ | Initial apical voiced stop/*d/ is palatalized to /*ɟ/ in Asmat-Kamrau Bay and devoiced to /*t/ in Muli Strait: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***d-** |***ɟ** |***t** | |body/contents|*dVnVmV |*ɟVnamV |*tɔnɔm | |hit/kill |*daɸa |*ɟaɸa- |*tap- | |sago palm |*dawɔ |*ɟawo | | |lie down |*dVi |*ɟ[a/o]i |*tai | Initial velar voiced stop /*g/ when followed by a front vowel /*i *e/ is retained as such in Asmat-Kamrau Bay and becomes phonetic [*j] in Muli Strait, where velar non-stop /*ɣ/(below) is not found folowed by front vowels: |< - 100px 100px 100px 100px >| | |Asmat-Muli |Asmat-Kamrau|Muli Strait| | |***g--/i e**|***g** |***j** | |give |*gema |*gema- |*je̝m- | |stab/kill |*geu- |*g[ew]u- |*je̝u- | |take/carry|*geua- |*gewa- |*j[i]u- | When followed by a non-front vowel, initial /*g/ is lenited to non-stop /*ɣ/ in Muil Strait and lost in Asmat-Kamrau Bay: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***g-/…** |***ø** |***ɣ** | |drum |*gaimV |*eme |*ɣaim | |2 sg. |*gV |*o | | |head/hair|*gVpV |*uɸu |*ɣo̝w | Initial bilabial non-stop /*w/ is occluded to voiceless stop /*p/ in Muli Strait, probably having been voiced /*b/ at an intermediate stage (see medial /*w/ below, also cf. Drabbe's 1949: 19 Ndom [wot] "garden"): |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***w-** |***w** |***p** | |garden/clearing|*wasa |*wasa |*par | Apical non-stop /*r/ has not been found initially.\\ \\ Initial palatal non-stop /*j/ is occluded to laminal voiced /*z/ in Muli Strait: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***j-** |***j** |***z** | |hear/sense|*jiu- |*jiwV- |*ziu- | |sun |*jau[a] |*jawu |*zaua | |call/speak|*jage- |*jage- |*zaɣ- | |shoot |*jVɾi- |*jVɾ[i]- |*zVri- | In one example, initial /*j/ is expected in Asmat but for unknown reasons does not appear: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***j[]-** |***ø** |***z** | |thorn/tooth|*[j]jgi |*igi |*ziɣ | Any consonant can occur medially.\\ \\ Medial bilabial nasal /*m/ is retained as such in both families: |< - 100px 100px 100px 100px >| | |Asmat-Muli |Asmat-Kamrau|Muli Strait| | |***-m-** |***m** |***m** | |causative|*-mɔ |*-mo |*-mɔp | |ashes |*[i/u]m[i/u]dV|*umid[i/u] |*[i/e̝]mund| |sago |*amVsa |*amosa | | |do.make |*ke-mV- |*e-mV- |*ke̝mV- | |give |*gema |*gema- |*je̝m- | Medial apical nasal /*n/ is retained as such in both families: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***-n-** |***n** |***n** | |thought |*mVnipi |*minipi |*me̝nip | |nose/tip |*mVnVgV |*m[e/a]n[e] |*mæne̝ɣ | |body/contents|*dVnVmV |*ɟVnamV |*tɔnɔm | Medial bilabial voiceless stop /*p/ …: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***-p-** |***p** |**** | |sago pulp|*kapimi |*apimi | | Medial bilabial voiceless /*ɸ/ is reflected as /*p/ in Muli Strait: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***-ɸ-** |***ɸ** |***p** | |ear |*iaɸVnV |*iaɸ[a/o]ne |*ie̝pær | |cold |*iVpV[kV] |*juɸoko |*iVp[u]i | |tail |*… |*e[p/ɸ]e |*p[ie̝] | |hit/kill|*daɸa |*ɟaɸa- |*tap- | Medial apical voiceless stop /*t/ … |< - 100px 100px 100px 100px >| | |Asmat-Muli |Asmat-Kamrau|Muli Strait| | |***-t-** |***s** |***t** | |saliva/spittle|*mVtapu |*masa[p/ɸ]u |*mitap | |knowledge |*mat[ɔ]m[ɔ]|*mas[a/o]mo | | Medial laminal voiceless /*s/ …: |< - 100px 100px 100px 100px >| ||Asmat-Muli|Asmat-Kamrau|Muli Strait| ||***-s-** |**** |**** | || | | | Medial velar voiceless stop /*k/ …: |< - 100px 100px 100px 100px >| ||Asmat-Muli|Asmat-Kamrau|Muli Strait| ||***-k-** |**** |**** | || | | | Medial bilabial voiced stop /*b/ …: |< - 100px 100px 100px 100px >| ||Asmat-Muli|Asmat-Kamrau|Muli Strait| ||***-b-** |**** |**** | || | | | Medial apical voiced stop /*d/ ir lenited to non-stop /*ɾ/ in Asmat-Kamrau Bay: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***-d-** |***ɾ** |***d** | |walk/go |*iad[e]- |*jaɾ[e]- |*iad- | |night/dark|*iVɣVdVmV |*juɾama |*iɔɣdɔm | |gall/bile |*idVgV |*jiɾi |*idVɣ | Medial velar voiced stop /*g/ is generally lenited to non-stop /*ɣ/ in Muli Strait and lost in Asmat-Kamrau Bay: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***-g-** |***ø** |***ɣ** | |see/look |*pVgVɾV- |*boɾa- |*pe̝ɣe̝r | |night/dark|*iVgVdVmV |*jurama |*iɔɣdɔm | |gall/bile |*idVgV |*jiɾi |*idVɣ | Medial bilabial non-stop /*w/ is occluded to voiced stop /*p/ in Muli Strait: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***-w-** |***w** |***b** | |dog |*iuwuɾi |*juwuɾi |*i[u]bui | |breast|*awVgV |*awo |*abuɣ | Medial apical non-stop /*ɾ/ is retained as such in both families: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***-ɾ-** |***ɾ** |***r** | |see/look|*pVgVɾV- |*boɾa- |*pe̝ɣe̝r- | |shoot |*jVɾi- |*jVɾ[i]- |*zVri- | Medial palatal non-stop /*j/ is occluded to apical /*r/ inMuli Strait, presumably via /*z/ (see initial /*j/ above) which otherwise does not occur medially : |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***-j-** |***j** |***r** | |urine |*aiji |*eji |*[a]ir | |new |*aji |*aji |*ari- | |thigh |*ɣaje |*a[j]e |*ɣ… | |banana|*ɣVjVw |*… |*ɣo̝ro̝b | … subfinal consonants …\\ \\ Subfinal bilabial nasal /*m/ is retained as such in both families: |< - 100px 100px 100px 100px >| | |Asmat-Muli |Asmat-Kamrau|Muli Strait| | |***-mV** |***mV** |***m** | |night/dark |*iVgVdVmV |*juɾama |*iɔɣdɔm | |louse |*amV |*amo |*am | |knowledge |*mat[ɔ]m[ɔ]|*mas[a/o]mo | | |body/contents|*dVnVmV |*ɟVnamV |*tɔnɔm | |sago pulp |*kapimi |*apimi | | |drum |*gaimV |*eme |*ɣaim | Subfinal apical naasal /*n/ is reflected as non-stop /*ɾ/ in Muli Strait: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***-nV** |***nV** |***r** | |ear |*iaɸVnV |*jaɸ[a/o]ne |*ie̝pær | |sand|*sini |*sini |*sir | There is one apparent counterexample in whcih subfinal /*n/ is retained as such in Muli Strait: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***-nV** |***nV** |***n** | |centipede|*an[i] |*an[i] |*an | Subfinal bilabial voiceless stop /*p/ is retained as such in both families: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***-pV** |***pV** |***.p** | |thought |*mVnipi |*minipi |*me̝nip | |saliva/spittle|*mVtapu |*masa[p/ɸ]u |*mitap | Subfinal bilabial voiceless /*ɸ/ …: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***-ɸV** |***ɸV** |***p** | |pig |*[o/u]ɸV |*oɸo |*up | |head/hair|*gVɸV |*uɸu |*ɣo̝p | Subfinal … /*F/ …: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***-FV** |***FV** |***p** | |unripe|*kɔFɔ |*oFo |*kɔp | Subfinal laminal voiceless /*s is fricated to /*s/ in Asmat-Kamrau Bay and reflected as apical non-stop /*r/ in Muli Strait (q.v. Voorhoeve 1980: 69.) It is probably not possible to distinguish this from subfinal /*t/, if one existed: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***-sV** |***sV** |***ɾ** | |blood |*[i/e]sV |*ese |*ir | |sago |*amVsa |*amosa | | |excrement |*asa |*asa |*ar | |name |*uase |*uwase |*ur | |red ochre |*uas[ae] |*wasae |*uar | |tooth/sharp |*sisV |*sisV |*-sir | |charcoal/black |*sosV |*sosV |*sɔr | |garden/clearing|*wasa |*wasa |*par | Subfinal velar voiceless stop /*k/ is found in several Amsat-Kamrau Bay forms with credible outcomparisons (cf. Cook River /*jakam/ "fruit", Kolopom/ *uak/ "egg") …: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***-kV** |***kV** |**** | |fruit/seed|*[e]ak[e] |*eake | | |egg |*[o]k[a] |*oka | | Subfinal apical voiced stop /*d/ … |< - 100px 100px 100px 100px >| | |Asmat-Muli |Asmat-Kamrau|Muli Strait| | |***-dV** |***dV** |***nd** | |ashes|*[i/u]m[i/u]dV|*umid[i/u] |*[i/e̝]mund| Subfinal velar voiced stop /*g/ when followed by front vowels /*i *e/ is retained as such in Asmat-Kamrau Bay and lenited to non-stop /*ɣ/ in Muli Strait: |< - 100px 100px 100px 100px >| | |Asmat-Muli |Asmat-Kamrau|Muli Strait| | |***-gV/_i e**|***gV** |***ɣ** | |thorn/tooth|*[j]igi |*igi |*ziɣ | |call/speak |*jage- |*jage- |*zaɣ- | When followed by non-front vowels, subfinal /*g/ is lost in Asmat-Kamrau Bay: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***-gV** |***V** |***ɣ** | |breast |*awVgV |*awo |*abuɣ | |nose/tip|*mVnVgV |*m[e/a]n[e] |*mæne̝ɣ | Subfinal bilabial non-stop /*w/ like initial and medial /*w/ (above) is occluded to voiced stop /*b/ in Muli Strait: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***-wV** |***wV** |***b** | |foot |*mawu |*mawu |*mɔb | |sago palm|*dawɔ |*ɟawo | | |crocodile|*kewV |*ewo | | Subfinal apical non-stop /*r/ is lost in Muli Strait: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***-ɾV** |***ɾV** |***ø** | |dog|*iuwuɾi |*juwuɾi |*i[u]pui | …vowels and diphthongs …\\ \\ As in neighboring Marind and Bulaka River, high vowels /*i *u/ in sequences with other vowels were distinguished from no-stops /*w *j/ (above) by the failure of the former to undergo occlusion in Muli Strait, a contrast which has been mostly levelled in Asmat-Kamoro: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***iV** |***jV** |***jV** | |ear |*iaɸVnV |*iaɸ[a/o]ne |*ie̝pær | |walk/go |*iad[e]- |*jaɾ[e]- |*iad- | |night/dark|*iVgVdVmV |*juɾama |*iɔɣdɔm | |cold |*iVpV[kV] |*juɸoko |*iVp[u]i | |dog |*iuwuɾi |*juwuɾi |*i[u]bui | …: |< - 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamrau|Muli Strait| | |***uV** |***wV** |***uV** | |name |*uase |*uwase |*ur | |red ochre |*uas[ae] |*wasae |*uar | |hear/sense|*jiuV- |*jiwV= |*ziu- | |sun |*jau[a] |*jawu |*zaua | ===== Pronouns ===== **[under construction]** ===== Verbal morphology ===== One of the most striking grammatical similarities between the two subfamilies are the forms of the postposed verbal subjects. As nothing is known about Kamrau Bay or Konorau verbal morphology, Asmat-Kamoro and Mombum (Drabbe 1950: 564-566) are compared directly: |< - 100px 100px 100px 100px 100px >| | |Asmat-Muli|Asmat-Kamoro|Mombum |Mombum | | | | |indicative|imperative| |1 sg.| | |-u |--- | |2 sg.| | |-im |-u | |3 sg.| | |-i |--- | |1 pl.| | |-am |--- | |2 pl.| | |-am |-mi | |3 pl.| | |-a/-i |--- |