Chapter 5.10 Cognition

Malcolm Ross and Meredith Osmond

1. Introduction⇫

A cognition verb like ‘know’, ‘think’, ‘understand’ or ‘remember’ denotes a concept that speakers are aware of because it denotes an event within their own minds, but often has only indirect correlates in the perceived world. As a result, speakers of different languages classify cognitive events in rather different ways, requiring us first to gain some insight into how speakers of present-day Oceanic languages classify these events.

English cognition verbs tend to cover a range of events. The verb think has a considerable range of meanings:

  1. Don’t talk to me—I’m thinking. (cogitation)
  2. I think John stole the key. (belief, opinion)
  3. I didn’t think of it (‘I forgot it.’)
  4. I thought I would go shopping (intention)
  5. I keep thinking about poor Mary (‘I’m worried because she is ill’ OR ‘I’m saddened by her death’ OR ‘I would like to be with her’)

To be sure, a native speaker disambiguates each meaning in context. The progressive aspect in the present tense (… am thinking) in (1) indicates that this is thinking in the sense of cogitation. The complement clause (… I would go shopping) in (4) points to intention.

There is probably no other language in the world with a verb whose range of meanings exactly corresponds to those of English think (not even close neighbours like French or German do), but many of our sources give English glosses consisting of a single cognition verb like ‘think’, leaving us ignorant of how the verb thus glossed is used.

To gain insight into how speakers of present-day Oceanic languages classify cognitive events, we have first tried to ensure that we compare like with like semantically. A list of semantic frames for cognition terms was drawn up. A semantic frame is a description of an event, relation, or entity and the participants involved in it.1 Making the list was a two-step procedure. First, the FrameNet website was consulted. It provides semantic frames for a very large number of English lexemes and, for example, distinguishes the various senses of English think. Second, frames were defined that reflect meanings found in dictionaries of Oceanic languages for cognitive states and activities. Semantic frame labels appear below in small capitals. Terms for each frame were found in dictionaries of four Oceanic languages: Nakanai (MM; Chowning and Goodenough 2014), To’aba’ita (SES; Lichtenberk 2008), Mwotlap (NCV; François 2012) and Wayan Fijian (Pawley & Sayaba 2003) and are tabulated in the sections on knowing (§10.2), thinking (§10.3) and remembering (§10.5). This constituted a check of the appropriateness of the list of frames and of their possible representation in POc. In the event, several cognition frames that were supported by dictionary glosses did not lead to the reconstruction either of forms or of metaphorical structures, and they are omitted here. These include ‘not know, be ignorant’ (often a simple verb), ‘think about, long for’, ‘be on one’s mind, have s.t. on one’s mind’, ‘remember to do s.t.’, ‘forget to do s.t.’, ‘hope’ and ‘expect’.

A larger language sample would have been ideal, but identifying semantic frames requires sentence examples. These are absent from Chowning and Goodenough (2014), but the latter is the best available dictionary of a MM language. Because semantic frames are subject to borrowing by bilingual speakers, and NNG and PT languages have all been in contact with Papuan languages at various points in their histories, they are probably poor indicators of POc’s frames and were therefore excluded from the sample, meaning that WOc could be appropriately represented only by a MM language.

There is a tendency for terms denoting abstractions to be metaphors that refer to less abstract concepts. Metaphors in turn are often encoded by complex lexemes; that is, lexemes made up of two or more simple lexemes. Complex lexemes include body-part metaphors (BPMs; ch.9), serial verb constructions (SVCs),2 and compounds derived from either of these, and apparently these have long been productive lexeme-creating devices, as they are also present in Central Malayo-Polynesian and South Halmahera/West New Guinea languages and were apparently constructions of Proto Central/Eastern Malayo-Polynesian. We can be sure that complex lexemes with these structures occurred in POc.

Each section below discusses a single cognition frame or a set of related frames. Sections discussing further frames could be added, but these would not contain reconstructed forms. They would at best list the meanings of complex lexemes together with supporting data, and these are already well enough represented in the chapter.

2. Knowing⇫

Verbs encoding three semantic frames denote knowledge in Oceanic languages:

__block - AWARE, e. g. ‘I know that he is coming.’ - ACQUAINTED, e. g. ‘I know him well.’ - EXPERT, e. g. ‘I know how to plant yams.’

Their distribution across verbs in the four witness languages is shown in Table 22.3 In Mwotlap, Wayan and To’aba’ita one verb is used for all three frames, but To’aba’ita also has dedicated EXPERT verbs. Nakanai has distinct verbs in each frame, but the AWARE verb rovi also occurs in the ACQUAINTED compound rovi-lala. The morpheme -lala is perhaps related (diachronically, at least) to lalai ‘to try (to do s.t.)’. If so, it has a similar meaning to To’aba’ita toʔo, which means ‘to try, test’ in a number of compound verbs (§8.5) including apparently θaitoʔoma- ‘know’, but does not occur independently.

Table 22 Verbs of knowing in the four witness languages
AWARE ACQUAINTED EXPERT
‘know (s.t. /that …)’ ‘know/recognise (s.o.)’ ‘know/learn (how to …)’
Nakanai rovi mata-kilala [look-(know)] tahai, mari
rovi-lala [know-? ]
To’aba’ita θai-toʔoma- [(know)-?] θai-toʔoma- [(know)-? ] θaitoʔoma-, filo-, filoŋani-, maʔalutani- [eye-? ], dau-fīfirisi- [? -thoroughly]
Mwotlap eɣlal eɣlal eɣlal
Wayan kilāti- kilāti- kilāti-

The glosses on the second line of Table 22 are intended to capture the fact that in certain contexts (e. g. in the presence of a perfective marker) ACQUAINTED and EXPERT verbs often have dynamic punctual senses, respectively ‘recognise (s.o.)’ and ‘learn (how to …)’.

The POc ‘know’ verb with the most widely distributed reflexes is *kilala. It appears to have had AWARE, ACQUAINTED and EXPERT senses, to judge from the more specific glosses in the cognate set below, but it is difficult to be certain. WOc glosses match the PMP gloss, ACQUAINTED. The trisyllabic form is unusual, and there is reasonable evidence for a transitive alternant *kila-i- from which the third root syllable was deleted.

PMP *kilala ‘know (a person), recognise, be acquainted with; feel, perceive’ (ACD)
POc *kilala
POc *kilala-i-, *kila-i- [VT] ‘know’
Adm Mussau kile ‘know’
NNG Lukep -kil- ‘recognise’
NNG Mangap -kilaala [VT] ‘know well, recognise, be aware, understand’
NNG Manam -kilala ‘recognise’
NNG Bariai kilala [N] ‘memorial, monument, mnemonic’
NNG Amara klele [VT] ‘know’
NNG Aria -ile [VT] ‘know (s.o.)’
MM Nakanai (mata)kilala ‘know, recognise (s.o.)’
MM Madak kilem ‘know’
MM Kubokota ɣila-ɣila ‘know’
MM Lungga ɣi-ɣila-i ‘know’
MM Nduke ɣi-ɣile- ‘know’
MM Roviana ɣilani- ‘know’ (-n- for †-l-)
MM Hoava ɣilali- ‘know’
SES Birao hila-hila ‘know’
SES Lengo ɣila-ɣila- ‘know’
SES To’aba’ita ʔilala ‘perform divination’
SES Arosi ʔirara ‘know, understand, perceive’
SES Owa ɣirara ‘know’
TM Natügu klʌ ‘know’
NCV North Ambrym kela ‘know’
NCV Paamese kilea ‘know, know how to, be able to’
NCV Lewo kilia [VT] ‘know, understand’
NCV Mota ɣilala ‘know, understand’
NCV Mwotlap eɣlal ‘know’
NCV Sakao köl ‘look for, find’
SV Sye okili ‘know’
NCal Nemi hina ‘know’
NCal Iaai xanā ‘know’
PMic *kila, *kila-a, *kila-i- ‘know’
Mic Kosraean (a)kile(n) [VT] ‘notice’
Mic Kiribati kinā ‘recognise, know’
Mic Kiribati kina-i [VT] ‘recognise, know’
Mic Marshallese kile-y ‘recognise, realise, distinguish, be familiar with, identify, notice, perceive’
Mic Chuukese siɾe ‘know how (to do s.t.), be skilled’
Mic Chuukese siɾe-e- [VT] ‘know s.o.’
Fij Bauan kila[-] ‘know, understand’

A number of languages have verbs that are formally similar to the reflexes above but have meanings that indicate that they more probably reflect POc *kilat (VI) ‘be seen clearly, discerned, recognised’, (VT) ‘see clearly, discern, recognise’ (§8.2).

NCV Tolomako kile- ‘see’
NCV Araki kila ‘watch, look (in a certain direction)’
NCV Atchin kila ‘look round, down’
NCV Avava kil-kila ‘look, open eyes’
Mic Ponapean kila(ŋ) ‘see, discern, look at, observe, examine’
Mic Woleaian xle ‘be clear, seen clearly, recognised’
Pn Rennellese kiga [VSt] ‘be clearly seen, in plain sight’

The Wayan verb kilāti- ‘know’, on the other hand, conflates a form reflecting *kilat with the sense ‘know’.

From the glosses of the data below, POc *qataq, *qataq-i- evidently meant ‘know, understand, realise (that)’, encoding AWARE. In a few languages the verb has the same form as the reflex of *qate- ‘liver’ (§3.7.6). Despite the role played by *qate- in bodypart metaphors, particularly those expressing emotions (§9.2.1), however, the resemblance seems to have emerged by chance. Final *-q is attested in Mutu and Namakir.

POc *qataq [VI] ‘know, understand, realise (that)’
POc *qataq-i- [VT] ‘know, understand, realise (that)’
Adm Nyindrou ata(na) ‘come to know, realise, understand’ (syntactically a verb, but the subject is encoded as a possessor suffix, e. g. atana-k ‘I realise’)
NNG Kilenge ota-i ‘know’
NNG Mutu wataɣ-i ‘know’
NNG Gitua wata ‘know’
NNG Bariai oata-i ‘know, learn’
NNG Kove ata-i ‘know’
NNG Mangseng ate ‘recognise, see that’
PT Iamalele ʔase(ta-i) ‘know, understand’
PT Dawawa kata-i ‘learn’
PT Tubetube kata-i ‘know’
PT Saliba kata-i ‘know’
PT Suau ʔata ‘know’
PT Misima ate(na) ‘know, understand’
PT Sudest ɣarei-ɣarei ‘know, understand’
MM Notsi ati ‘know’
MM Nehan ate, iate ‘know’
MM Halia atei ‘know’
MM Mono-Alu atae ‘know s.o.’
SES Longgu ðai- ‘know, understand, be accustomed (to doing); be able (to do)’
SES Marau Sound rae- ‘know’
SES Lau sai(toma), sai(tama) ‘know (s.t., s.o.)’
SES To’aba’ita θai(toʔoma-) ‘know’
SES ’Are’are rai- ‘know, understand’
SES ’Are’are rai hitari- ‘know well’ (hitari- (VT) ‘split’)
TM Asuboa kata ‘know’
NCV Namakir ʔataʔ ‘know’
NCV Nguna atae ‘know’
NCV Lelepa tae- ‘know’
NCV South Efate tae ‘know’
NCV South Efate (nroŋ)tae ‘recognise by hearing’
NCV South Efate (mro)tae ‘understand’ (mro ‘think’)
NCV South Efate (le)tae ‘realise, recognise, identify’ (le ‘look, see’)
PMic *ata, *ata-i- ‘know, understand’
Mic Ponapean ɛsɛ ‘know, understand (s.t.)’
Mic Kiribati ata-i [VI] ‘know, have knowledge’
Mic Kiribati ata-a [VT] ‘know (s.t.)’
Mic Kosraean etʌ ‘know, understand (s.t.)’

The first morpheme of PPn *qata-mai ‘intelligent, expert, clever’ evidently reflected POc *qataq.

PPn *qata-mai ‘intelligent, expert, clever’ (POLLEX)
Pn Tongan ʔatamai ‘intelligent, intelligence’
Pn Samoan atamai ‘intelligent, clever’
Pn Anutan atamai ‘mind, meaning’
Pn Tuvalu atamai ‘skilful, able; skill, ability’
Pn Emae atamai ‘wise, wisdom’
Pn Nukuoro adamai ‘recollect/recall past events/persons’
Pn Pukapukan atamai ‘wish, desire; intelligent, having common sense’
Pn West Futunan atamai ‘right-minded, sane, clever’
Pn Tahitian atama ‘wisdom, intelligence, wise, intelligent’
Pn Hawaiian akamai ‘clever, expert’
Pn Māori atamai ‘knowing, quick-witted; malicious’

POc evidently had another term with an EXPERT meaning, *taqu, but it is reflected with reasonable certainty only in Anejom (SV) and in Polynesian languages, and two PPn terms are reconstructable: *tau ‘skilful at, familiar with’ and *mātau ‘know, understand, be experienced’. The latter has an apparent Banoni (MM) cognate, allowing the reconstruction of POc *ma-taqu (*ma- was a stative formative; §1.3.5.4).

PAn *Caqu ‘know how, be able to, be skilled at’ (ACD)
PMP *taqu ‘know how, be able to, be skilled at’ (ACD)
POc *taqu ‘know how, be able to, be skilled at’
SV Anejom̃ a-tou ‘know, know how to, be able, understand, be certain, be sure’ (John Lynch, pers. comm.)
PPn *tau ‘skilful at, familiar with’ (POLLEX)
Pn Tongan tau ‘skill that one is accustomed to do’
Pn Tuvalu tau ‘proper, necessary, possible, compulsory’
Pn Pukapukan tau ‘to fit, look nice’
Pn Rarotongan tau ‘be suitable, befit, able, to be possible’
Pn Sikaiana tau ‘be fit or suitable’
Pn Takuu tau ‘equal to a task’
Pn Tikopia tau ‘be accustomed, used to, adapt, fit’
Pn West Futunan tau ‘follow in the ways of, take after, learn from’
Pn Māori tau ‘be able, be suitable’
cf. also:
NNG Manam to ‘learn’
POc *ma-taqu ‘know, understand, be experienced’ (also ‘right-hand’: §3.6.3)
MM Banoni matō ‘know, be smart’
Fij Wayan mātau [VSt] ‘be familiar to s.o’ (subject the thing that is familiar); ‘accustomed to, used to’ (experiencer marked by oblique case)
Fij Wayan matau ‘right-hand side’
Fij Bauan matau ‘be right-handed’
PPn *mātau ‘know, understand, be experienced’ (POLLEX)
Pn Tuvalu matau ‘clever, experienced, right hand’
Pn Tongarevan mātau ‘accustomed to, usual’
Pn Rarotongan mātau ‘have knowledge of, be accustomed to, be in the habit of’
Pn Tuamotuan mātau ‘understand’
Pn Māori mātau ‘know, understand’
cf. also:
Fij Rotuman macau ‘be expert, skilful’ (-j- for †-t- or †-f-)

It is well known that in many languages a perception verb may also mean ‘know, understand (s.t.)’ (Aikhenvald & Storch 2013, Evans & Wilkins 2000, Viberg 1984). English uses ‘I see’ to mean ‘I understand’, i.e. an AWARE sense. This semantic extension occurs occasionally in Oceanic languages. A few NNG languages use a reflex of POc *reki[-], *reqi[-] ‘see, look, see s.t., look at s.t.’ (§8.2) also in the sense ‘know’:

NNG Mangap re ‘see, look, experience; consider, think, be aware’
NNG Yabem liʔ ‘see, look at s.t., know, have experience’
NNG Lamogai rik ‘see, know’

A similar extension of meaning occurs with PPn *kite ‘see, appear, know’ from POc kita-i- ‘see s.t.’, and Raga (NCV) ilo ‘know, perceive’ from POc *qilo ‘be aware of, discern, see’ (§8.2). The transitive reflex of POc *qilo, namely PPn *qilo- (VI) ‘to know, be aware’, (VT) ‘know s.t.’, had been fully repurposed as a verb of knowing.

Reflexes of POc *roŋoR- ‘hear s.t., listen to s.t.’4 with the additional sense ‘know’ are sufficiently widespread to raise the possibility that this sense was already present in POc.

NNG Mutu -lōŋ ‘know how to’
NNG Bing -luoŋ ‘know’
NNG Takia -loŋ ‘hear, listen, perceive, know’
NNG Gedaged -loŋ ‘know, have knowledge of, be aware of, hear, learn, perceive, understand’
PT Wedau -nonori ‘know’
PT Mekeo loŋo ‘know’
PT Northwest Mekeo oŋo ‘know’
MM Nakanai lolo ‘hear, understand, know’
MM Sursurunga a-loŋr-a ‘hear; listen and understand’
MM Nehan loŋoro ‘hear, understand’
SES Sa’a roŋo ‘hear, listen, hear tidings of, understand’
NCV Lakon ruŋ ‘hear, feel; obey, know’

Lexical replacement has evidently been frequent among verbs of knowing, and many reconstructions can be made of verbs that are reflected in just one subgroup. Some are listed here in the hope that their origins may eventually be identified.

Proto Willaumez *maci ‘know’ (Goodenough 1997)
MM Bola mari ‘know’
MM Bileki mari ‘know’
MM Maututu masi ‘know’
PPT *siba ‘know’
PT Bohutu siba ‘know’
PT Hula riba ‘know’
PT Balawaia riba ‘know’
PT Motu diba ‘know’

The verbs below reflect *sagova, *sagov-i- ‘know’, reconstructable to a lower-order interstage within the Papuan Tip cluster.

PT Gumawana -yagoi- ‘understand s.t., know s.t. /s.o.’
PT Iduna -yakovi- ‘recognise s.o.’
PT Gapapaiwa -akova [VI] ‘know, understand’
PT Anuki -akovi- ‘know’
PT Ubir -sagob ‘know’

All languages below reflect *b⟨in⟩isi, but Nokuku also reflects *bisi, implying that *bisi is the root and that *⟨in⟩ reflects the POc nominalising infix, the resulting nominalisation having been reanalysed as a verb in these languages.

PNCV *bisi, *b⟨in⟩isi ‘know’
NCV Raga binihi ‘think, consider’
NCV Nokuku pi-pinis ‘know’
NCV Nokuku pisi- ‘know, understand, be able’
NCV Tolomako pinisi- ‘know’
NCV Kiai pinisi [VI] ‘be able to, know’
NCV Kiai pinisi- [VT] ‘know’

Interestingly, many Oceanic languages have distinct verbs for ‘not know (s.t.), be ignorant of (s.t.)’ and for ‘not recognise (s.o.)’, but none of the terms found is cognate with any of the others. Some terms are evidently monomorphemic, like Lou tɔn ‘not know’, Mangap -kus ‘not recognise’, Takia -ŋaoŋ ‘not know’, whilst others, like Balawaia ɣita-lea ‘not recognise’ (ɣita ‘see’ + lea ‘miss’) and Wayan kila sēti- ‘not recognise’ (kila ‘know (s.t., s.o)’ + sēti- ‘do s.t. wrongly’) are clearly serial verb constructions.

3. Thinking⇫

Across languages verbs of thinking fall into two broad semantic frames:

  • OPINE, e. g. ‘I think/believe that he is coming.’
  • COGITATE, e. g. ‘I think of/about him/this a lot.’

Table 23 shows that in all four witness languages there is a verb (in bold) that embraces both frames, but in Nakanai, To’aba’ita and Wayan there are other verbs with somewhat more specialised meanings. None of this is surprising. English has believe, surmise, guess, suspect and suppose as OPINE verbs, and a number of COGITATE verbs: cogitate on, consider, ponder, reflect on, contemplate and others, each with a different shade of meaning. Dictionaries often do not encapsulate these shades of meaning well.

Table 23 Verbs of thinking in the four witness languages
OPINE COGITATE
‘think/believe (s.t. /that …)’ ‘think about/consider (s.t.)’
Nakanai gabu, ule, vei, kau gabu, aliale, loho-tavu [cogitate-towards], ilo-tavu [inside-towards], hilo-tavu [see-towards]
To’aba’ita manata, sore- manata-i-, loloma, ono-toʔo- [belly-(test)]
Mwotlap dem dem
Wayan nūmi- nūmi-, lēŋa-i-

OPINE verbs seem to occur less frequently in Oceanic discourse than they do in European languages, and there are at least two reasons for this.

First, OPINE verbs differ from AWARE verbs (§10.2) in that a complement clause of the latter is taken to be a fact, whereas the complement clause of an OPINE is not. I know John is a teacher entails the proposition John is a teacher as a fact, but I think John is a teacher doesn’t guarantee the truth of the proposition. One result of this is that in English I think is often little more than a marker of possibility, i.e. ‘perhaps’. The Tok Pisin term for ‘perhaps’ is ating, transparently reflecting English I think, and many Oceanic languages have a corresponding sentence adverb that is glossed in dictionaries ‘perhaps, I think’; e. g., Mangseng (NNG) ava, Misima (PT) tabam, Muyuw (PT) adók, Tawala (PT) nugote, Ramoaaina (MM) bi-gaŋ, Sursurunga (MM) gut, Teop (MM) aekas, Kwaio (SES) baleʔe, Mwotlap (NCV) so. Of these, however, only the Tawala adverb has a derivational relationship to an OPINE verb (see below), and it seems that in Oceanic languages OPINE verbs typically do not have this bleached ‘perhaps’ function.

Second, OPINE is quite often expressed by a languages’s default verb of saying, so that in Baluan (Adm), for example, it is sometimes difficult to tell whether the speaker intends the complement of pʷa to be spoken or simply thought (Dineke Schokkin, pers. comm.). Bugenhagen & Bugenhagen (2007b) gloss the Mangap sentence

as both I say that is not good and I think that is not good. Thus the meaning of the example ‘say/think’ verbs listed below is something like ‘formulate in words, either spoken or unspoken’.

Adm Baluan pʷa [VT] ‘say, express, think’
Adm Nyindrou aña ‘think, say’
NNG Bariai oaŋga ‘think, say’
NNG Kaulong vo ‘talk, say, speak; suppose, intend’
NNG Mangap -so ‘say, speak, communicate, talk, tell; think’
NNG Takia -bol ‘say, talk, speak,’
PT Iamalele vo ‘say, think’ (quotative marker)
MM Nakanai vei ‘think, opine, talk, tell say’
MM Teop boha ‘think, say’
SES Gela ne ‘say, think’
SES To’aba’ita sore- ‘say, think’
SES Kwaio ilia ‘say, tell, think’

Hence OPINE verbs in Oceanic languages are centrally about mental activity, and it is not surprising that Table 23 shows them overlapping with COGITATE verbs. Indeed, no dedicated POc OPINE verb that is not also a speech verb is reconstructable.

Glosses of reflexes of POc *nonom ‘think, remember; mind, thought’ point quite strongly to it being a COGITATE verb with a semantic focus on thinking about or remembering something. Its reconstruction, though, involves some ad hoc assumptions about the history of the apparent reflexes listed below. These display a somewhat abstract formal template nVNV[N], where N is n or m, but m occurs no more than once in a reflex. The shape is that of POc *nonom (V) ‘think’, (N) ‘mind, thought’, the expected reflex of PAn *nemnem ‘think’ (ACD). However, Blust (ACD) notes Fordata (CMP) nanaŋ ‘remember, remember sadly’, with -a- twice for expected -e- (< PAn *-e-). This suggests that there was an alternant of the form *nanam as far back as PCEMP, perhaps ancestral to some of the forms listed below. Treating the forms below as a cognate set also assumes that the presence of three nasals led to metathesis in Seimat and Nehan (*nVnVm > *nVmVn), and to assimilation of point of articulation in Bariai, Babatana and Ririo (*nVnVm > *nVnVn).

The Wayan transitive verb num-i- (VT) ‘think of s.t.’ requires special mention. As Blust (1977a) shows, a POc intransitive verb of the form C1V1-C1V1C2 often had a corresponding transitive of the form C1V1C2-i. Thus POc *nonom may have been paired with transitive *nom-i, of which Wayan num-i- is the only reflex known to us. Alternatively, it may be a back-formation from intransitive *nanum, reflected in Bauan nanu.

PAn *nemnem ‘think’ (ACD)
POc *nonom, *nanam ‘think about s.t., remember s.t.’; [N] ‘mind, thought’
Adm Seimat namena [VI] ‘remember’ (metathesis of nasals)
NNG Bariai nanan ‘think, remember’
PT Kilivila nano ‘mind’
MM Nehan namana ‘think; think about s.t.’ (metathesis of nasals)
MM Babatana nanana [V] ‘think’; [N] ‘thought, mind’
MM Ririo (no)nono ‘think’
MM Roviana nonoŋa ‘remember, know’
NCV Mwotlap nonom ‘opinion’
NCV Southeast Ambrym nɛnem-i- ‘think, remember’
NCal Nêlêmwa nanam ‘thought, think, reflect, believe’
Fij Wayan num-i- [VT] ‘think of s.t.’
Fij Bauan nanu [VI] ‘think, meditate, remember’
Fij Bauan nanum-a [VT] ‘think of, meditate on, remember s.t.’
cf. also:
NNG Lukep nan(tut) ‘remind’
NNG Mengen nan(guni) ‘think, surmise’
MM Babatana nüni ‘think’
MM Vaghua nanavu ‘think’
MM Varisi nanao [V] ‘think, consider’; [N] ‘idea’

The two verbs discussed below, POc *drodrom ‘think, worry; love, be sorry for, long for’ and POc *nuka ‘think, feel’, are both COGITATE verbs, but both also have emotional overtones. Indeed, glosses in Oceanic dictionaries suggest that cogitation and worry or longing frequently go together.

Although its reflex is the Mwotlap default verb for thinking (Table 23), the NCV evidence, presented in some detail below, suggests that the POc verb was a COGITATE verb with an emotional overtone of ‘love, be sorry for, long for’, i.e. the SORRY semantic frame recognised in §11.4.3. Indeed, the emotion-related meanings are the only ones recorded for the Nakanai, Nokuku, Namakir, Nguna and S Efate reflexes, and they also figure in the Tamambo and Uripiv glosses. The expected POc reflex of PAn *demdem is POc †*rodrom (*-md- > *-nd- > -dr-), but maintaining the consistency of reduplication is perhaps responsible for *drodrom. Transitive *drom-i arose via the template recognised by Blust (1977a).

PAn *demdem ‘brood, hold a grudge, remember, keep still’ (ACD)
POc *drodrom [VI] ‘think, worry; love, be sorry for, long for’
POc *drom-i [VT] ‘think, worry; love, be sorry for, long for’
NNG Mangseng (lemi-) rum ‘think’ (lemi- ‘insides’)
NNG Mangseng rum(oŋ) [N] ‘thought’ (-oŋ NOMINALISER)
NNG Mengen roma ‘think about’
MM Nakanai gogo ‘be sorry for, be fond of, treat gently; be generous to’
MM Madak doma [V] ‘think’
PNCV *dodomi ‘think about, love’ (Clark 2009)
Proto Torres-Banks *do-domi ‘think, worry’ (François 2005)
NCV Dorig dum ‘think, worry’
NCV Nume dudum ‘think, worry’
NCV Mwesen nunum ‘think, worry’
NCV Mota nom ‘think, have in mind’
NCV Mota no-nom ‘think’
NCV Mwotlap dem ‘think’
NCV Nokuku ʔomi ‘love, have mercy on’
NCV Kiai komi-a [VT] ‘think of’
NCV Kiai komi-komi ‘thinking, thought’
NCV Tamambo domi ‘feel sad about, sorry’
NCV Tamambo domi-domi ‘think’
NCV Sakao rem [VI] ‘think’
NCV Sakao röm [VT] ‘think’
NCV Ambae domi ‘think’
NCV Uripiv (o)r̃m-i ‘think, worry, regret, have pity, show mercy’
NCV Labo rur(uox) ‘think’ (uox ‘follow’)
NCV Lonwolwol dɛmɛ ‘think’
NCV Paamese demi ‘think, believe; think about’
NCV Namakir do-dom ‘love, feel emotion’
NCV Namakir do-do- ‘mind’
NCV Nguna do-domi-a ‘love, be sorry for, feel for, miss’
NCV South Efate ⁿrom [V] ‘love’

Blust (ACD) reconstructs PAn *ajem ‘heart, mind’. Reflexes are found in SE Solomonic languages, some of which reflect an unexpected initial *q-.

PAn *ajem ‘heart, mind’ (ACD)
POc *(q)ajom [(VI) ] ‘think, understand’
POc *(q)ajom-akin-i- [(VT) ] ‘think, understand’
SES Gela ado-ado ‘think, understand’
SES Sa’a adom-aʔini ‘think’
SES Arosi ʔado-ʔado ‘think’
SES Arosi ʔadom-aʔi ‘think’
SES Fagani kato-katom-aɣi ‘think’

POc *nuka ‘think, feel’ was also evidently a COGITATE verb, but with a sense of associated emotion—desire for its object. In some daughter languages the reflex of *nuka is a verb, in others a monovalent body-part noun meaning ‘mind’, ‘thought’, ‘feeling’, or ‘desire’, and in yet others both a nominal and a verbal reflex occurs. When it occurs in complex lexemes, it is sometimes difficult to tell whether it is a verb or a noun, and a rule of thumb is adopted such that it is glossed as a verb ‘think’ unless there is clear evidence that it is a monovalent noun.

The reconstruction of *nuka is a little problematic with regard to its medial *-k-, and it is tempting to avoid irregularity by splitting the data into two formally similar cognate sets. However, the glosses imply quite strongly that this is a single set. The irregularity occurs in the Micronesian reflexes. Woleaian nʉ-nʉwa-n and Ifaluk nu-nuwa-n reflect either *nua or *nuqa, whereas Carolinian lɨxɨ-lɨx reflects *nuka. The Adzera medial -g- and Tolai and Ramoaaina final -k reduce the choice to *nuqa or *nuka, but could reflect either (final *-q is occasionally retained in New Ireland languages). Since *q is lost in Micronesian languages and the reflexes of *-k- in Chuukic languages like Woleaian and Ifaluk are known to be complex and not always regular (Jackson 1983:175–185), it makes sense to treat the Carolinian reflex as criterial and to reconstruct *nuka. The MM and PT reflexes in which *-k- is thus deemed to be lost are all regular.

POc *nuka [V] ‘think, feel’
POc *nuka- [N] ‘mind, thought’
NNG Adzera nugu- ‘insides, heart, seat of emotions’
PT Gumawana nue [VT] ‘think of s.t.’ (-nue < *nuka-i-)
PT Gumawana nuo-nuo- [N] ‘thinking, thoughts about s.t.’
PT Iduna -nua-nua [VI] ‘think’
PT Iduna -nua-nue- [VT] ‘think (about s.t.)’ (-nue < *nuka-i-)
PT Iduna nua-nua [N] ‘thought, desire, idea’
PT Bwaidoga nua- ‘mind, insides’
PT Gapapaiwa nua ‘feel, think’
PT Gapapaiwa nua-nua ‘feelings, thoughts’
PT Minaveha nua- ‘feelings, desire, thought’
PT Minaveha nua-nua- ‘knowledge, memory, desire’
PT Dobu nua- ‘mind, desire, thought, will’
PT Molima nua-nua ‘think’
PT Wedau nua-nua- [N] ‘chest; seat of the emotions’
PT Tawala nugo ‘mind’
PT Bunama nua-nua [V] ‘think, want’; [N] ‘mind’
PT Saliba nua ‘mind’
PT Muyuw nua- ‘abdomen, belly; insides’
PT Sudest (re)nua(ŋa) [N] ‘thought, mind’
MM Kara nə- [N] ‘thought, idea’
MM Madak nua ‘think’
MM Tolai nuk[-nuki] ‘mind, heart, soul, seat of thoughts or ideas’
MM Ramoaaina nuk [VT] ‘think, remember’
MM Ramoaaina nu-nuk [VI] ‘think’
Mic Carolinian lɨxɨ-lɨx ‘believe, think’
Mic Woleaian nʉ-nʉa-n [VI] ‘to think, remember’
Mic Woleaian nʉ-nʉa-n- [VT] ‘remember s.t.’
Mic Woleaian nu-nua-n [N] ‘thought, emotion’ (Lutz 1988)

The inherited core meaning of POc *manaca(m) was evidently ‘tame (of animals), familiar to’ (of people). Its form—*ma- + disyllabic root—indicates that it was originally a stative verb, but the glosses of the forms below suggest that it came also to be used of people in the senses ‘quiet, thoughtful, learned’, and then developed the meanings ‘know, understand, think about’ and was also used as an abstract noun. In a number of languages it became the base for a transitive verb. In some languages the original meaning has been lost, but the retention of ‘tame’ as one of its senses in Lau, ’Are’are, Sa’a, Arosi and Owa attests to something like this series of semantic developments. Reflexes vary in meaning between AWARE and COGITATE.

PAn *ma-Lajam ‘tame, accustomed to’ (ACD)
PMP *ma-najam ‘tame, accustomed to’
POc *ma-nacam [VI] ‘tame, quiet, thoughtful, learned; know, understand, think about’; [N] ‘knowledge, understanding, thought, wisdom’
NNG Gedaged mana-n ‘tame, docile (mostly of animals), peaceful, obedient, trained’ (for †manaya-n)
PT Motu manada ‘even, smooth, gentle’
MM Ramoaaina manā(na) [VI] ‘know, understand’; [N] ‘knowledge, understanding, wisdom’
MM Nehan mahanama ‘tame, unafraid’ (metathesis)
SES Gela manaha [VT] ‘know, understand, appreciate; wise, clever’
SES Lengo manaθa [N] ‘knowledge’
SES To’aba’ita manata [VI] ‘think’; [N] ‘thought, mind’
SES To’aba’ita manata-i- [VT] ‘think of, about s.t., think (that…)’
SES To’aba’ita manatā ‘thought, idea’
SES Lau manata [V] ‘tame, quiet, civilised, sensible, understanding, think, thoughtful, careful’
SES Lau manata- [N] ‘mind, will, understanding’
SES Lau manata-ŋa, manatā [N] ‘thought’
SES Kwaio manata ‘think, reason, know’
SES Kwaio manate-ʔe wane ‘a man’s mind’
SES ’Are’are manata ‘be tame (of birds and animals), behave oneself, wise, sensible, learned’
SES ’Are’are manata-na [N] ‘disposition, character, nature, custom, behaviour, conduct, knowledge, wisdom’
SES ’Are’are manata-ʔini- [VT] ‘know, be aware of, notice’
SES Sa’a manata [VI] ‘tamed, quiet, taught’
SES Sa’a manata-ŋa [N] ‘wisdom, nature, knowledge’
SES Arosi manata ‘tame, trained, gentle (of man or animal)’
SES Arosi manata-si- ‘be tame towards’
SES Arosi manata-na [N] ‘custom, use’
SES Owa manata ‘be tame; be familiar to’
SES Owa manata-si- [VT] ‘know (s.o.)’
cf. also:
MM Nehan manate ‘know’ (-t- for †-h-)

4. True and believing to be true⇫

In those Oceanic languages for which there are relevant data5 believing something to be true usually differs lexically from OPINE (§10.3) and thus forms a separate semantic frame, here labelled BELIEVE. In most of these languages, the basic BELIEVE predicate is a complex form, either a derived verb or, less commonly, a BPM, involving a stative verb root meaning ‘true, real, genuine, correct, right’, a frame here labelled TRUE. The most widespread derivation is a TRUE verb preceded by the prefix that forms causative verbs, reflecting POc *pa[ka]-. Verbs with this form are listed in Table 24.

From the examples in Table 24 it seems likely that there was a POc believe verb of the form *pa[ka]- + true verb, but its form is uncertain. The glosses of *pa[ka]- + true verbs in the table point to the likelihood that the basic meaning of POc *pa[ka]- + true was ‘verify as true’, and that ‘believe (s.t.) to be true’ was a secondary meaning. Other derivations with a true root are listed in Table 25. The Takia lexeme is a BPM, and the Owa lexeme is a compound derived from a BPM. The Gela, Longgu, Sa’a and Pn forms are evidently compounds derived from SVCs.

It follows from the material in Tables 21 and 22 that the term to be reconstructed is the stative verb for the true frame rather than a believe verb. In other words, this is an instance where the basic lexeme was a stative verb with the stimulus as subject: ‘X is true’ rather than ‘I believe X’.

Table 24 BELIEVE verbs formed from the causative prefix + a TRUE verb
BELIEVE TRUE
‘believe (s.t.) to be true’ ‘true, real, genuine, correct, right’
PT: Balawaia vaɣa-moɣoni ‘believe, agree, confirm’ moɣoni
MM: Teop va-mana-mana ‘believe’ mana
MM: Banoni va-cū ‘believe’ cu
MM: Babatana va-tuna ‘believe’ tuna
MM: Roviana va-hinokar-i- ‘believe; prove’ hinokara-
MM: Maringe fa-tu-tuani ‘believe’ tuani
SES: Bugotu va-utu-utuni ‘believe’ utuni
SES: To’aba’ita faʔa-mamana (VI) ‘be truthful, reveal the truth’ mamana
faʔa-mamane- (VT) ‘believe, give credence to’
SES: Arosi haʔa-momori ‘believe’ momori
Fij: Bauan vaka-dina-dina ‘confirm, witness’ dina
vaka-dina-t- ‘believe’
Fij: Wayan vaka-dū-ni- ’ ‘believe; confirm truth or accuracy of s.t.’ dū
Pn: Tongan faka-moʔoni ‘bear witness, prove, verify’ moʔoni
Pn: Niuean faka-mooli ‘witness, tell truth, prove’ mooli
Pn: Rennellese haka-māʔogi ‘verify as true’ māʔogi
Pn: Maori ɸaka-pono ‘believe’ pono ‘true; bountiful, abundant’

Table 25 Other BELIEVE lexemes formed with a TRUE verb
BELIEVE TRUE
‘believe (s.t.) to be true’ ‘true, real, genuine, correct, right’
NNG: Takia _ilo-rumok (ilo- ‘insides’) rumok ‘truth’
SES: Gela talu-utuni (talu ‘put’) utuni
SES: Tolo t-utuni- utuni
SES: Longgu naʔi-utuni (naʔi ‘put’) utuni (borrowed from a Guadalcanal language)
SES: Sa’a hī-walaʔimoli (hī ‘perceive’) walaʔimoli
SES: Owa raro-ni-mʷora (raro ‘insides’) mʷora
Pn: Samoan tali-tonu (tali ‘accept’) tonu ‘correct’
Pn: Tokelauan tali-tonu (tali ‘accept’) tonu ‘correct’

The most widely reflected TRUE verb is POc *tuna (sometimes *tutuna) ‘true, genuine, correct’.

POc *[tu]tuna ‘true, able to be believed, correct’
NNG Lukep tun ‘correct’
PT Misima tuna(hot) ‘that’s true; yes’ (hot emphatic)
MM Patpatar tun ‘correct’
MM Patpatar tu-tun ‘true, faithful, responsible, real in form or appearance’
MM Ramoaaina (liŋ)ta-tuna ‘true; truth’
MM Tolai tuna ‘real, true, proper, correct’
MM Babatana tuna ‘true, real’
MM Babatana tu-tuna ‘true, just; truth’
MM Babatana (va)tuna ‘believe’
SES Arosi (hu)una ‘real, true, original’
cf. also:
Adm Lou tuɛna- ‘true’ (origin of -ɛ- is unknown)
Fij Bauan dina ‘true; very’ (-i- for †-u-)
Fij Bauan (vaka)dina-t- ‘believe’

It is tempting to combine the set below with the set above. All the forms above could reflect putative *tuquna, with regular loss of *-q- and shortening of resulting *-uu-. However, none of the forms below would be regular reflexes, as they fail to reflect either *-a or *-na as predicted by regular sound change. Either the formal similarity between *tuna and *(t,d)uqu is accidental, or they were associated at some point in their history by an unknown derivational process.

POc *(t,d)uqu ‘true, able to be believed’
Adm Nyindrou (ha)dru ‘true; very, really’ (reflects *d-)
NNG Dami tu-tuk ‘correct, innocent’
MM Banoni cu ‘true’
MM Banoni (va)cū ‘believe’
NCal Cèmuhî ju, jū ‘true’ (reflects *d-)
Fij Wayan dū ‘right, correct, genuine, real, true’ (reflects *d-)
Fij Wayan (vaka)dū-ni- ‘believe’

Overlapping semantically with the TRUE frame is the STRAIGHT frame, as Oceanic verbs meaning ‘straight’ tend strongly also to have the metaphorical sense ‘correct’, a component of the TRUE frame. Some reflexes of POc *tonuq ‘straight, correct’ have the additional sense ‘true’, and it seems possible that contamination by reflexes of *tuna has occurred, resulting in forms that appear to reflect †*tunuq rather than *tonuq. On the strength of Nokuku ta-tino ‘true’ and Kiai tu-tunu ‘good, straight, sweet’ below, all the NCV forms have been attributed to *tunuq, but some may either reflect *tuna above or a contamination of one form by the other.

POc *tonuq ‘straight, correct’6
NNG Bam tun-tunu ‘straight’
NNG Numbami tonowa ‘straight’
PT Minaveha tunuɣa ‘straight’
PT Minaveha tunu-tunuɣa ‘do right, be righteous’
PT Iduna tunu-tunuɣ(ina) ‘straight (of objects, path), upright, honest’
PT Molima tunu-tunv(ina) ‘straight, flat’
MM Laghu to-tonu ‘straight’
NCV Nokuku ta-tino ‘true’
NCV Kiai tu-tunu ‘good, straight, sweet’
NCV Uripiv (were)tun ‘tell truth’
NCV West Ambrym ten ‘real’
NCV Southeast Ambrym (rei)tin ‘true’
NCV Lonwolwol ten ‘real’
NCV Lonwolwol (fɪ)tɛn ‘true; truly’
NCV North Ambrym (fɛ)tɪn ‘true; truly’
NCV Paamese tine ‘true’
NCV Avava (ba)rīn ‘true’
PPn *tonu ‘straight, correct’ (POLLEX)
Pn Tongan tonu ‘exact, correct, be right’
Pn Niuean tonu ‘proper, right’
Pn Samoan tonu ‘exact, correct, just’
Pn Tuvalu tonu ‘straight, correct’
Pn Ifira-Mele tō-tonu ‘right, correct’
Pn Tikopia tonu ‘right, correct, true, exact’
cf. also:
NNG Mangap du-dūŋ ‘real, correct, straight’
NNG Malai dunu(ŋa) ‘straight’
SES Bugotu jino ‘straight, right, righteous’ (-i- for †-u-)
NCV Mota nun ‘true, truth’
SV Kwamera a-tuən ‘implies straightening’ (verbal adjunct)
Fij Bauan donu ‘straight, correct, true’
Fij Wayan donu ‘right, correct, true’

Several forms with initial *m- meaning ‘true’ can be reconstructed. The reason is perhaps that each has its origins in a form with the PMP anticausative/stative prefix *ma-. This is certainly true of reflexes of POc *ma-qoli and *ma-qoni, both ‘true, real’. Despite their formal and semantic similarity, they appear to have been separate POc terms. Their similarity has almost certainly led to crossovers in meaning and to conflation of the two terms, as apparently no language other than Anutan reflects both—and the gloss of Anutan maori ‘indigenous, true, close of kin’ suggests it is borrowed from an EPn language. No EPn language has a reflex of *ma-qoni. PEPn *ma-qoni acquired the additional sense ‘native, indigenous’, giving rise to the terms Māori and Mōriori for the Polynesian inhabitants of New Zealand and New Zealand’s Chatham Islands respectively.

FIXME: erratum inn vol2

Perhaps the clearest indicator that the terms originally had slightly different meanings is the contrast in meaning between the PPn causatives PPn *faka-moqoli ‘assent (V)’ and PPn *faka-maqoni ‘tell the truth, be honest’.

There is evidence that Gela, Lau and S Efate reflexes (shown under ‘cf. also’ below) of POc *ma-qoli ‘true, real’ have been conflated with those of POc *maqurip ‘be alive, live, flourish’ (§4.2.1.1). All three reflect POc *-r- rather than *-l-, and the Gela and Lau reflexes mean ‘alive’ as well as ‘real’.

POc *ma-qoli ‘true, able to be believed’
MM Bola muɣoli ‘true’
SES ’Are’are (wara-ʔi)mori ‘true’ (wara ‘speech’)
SES Arosi mori, mo-mori ‘true’
SES Arosi haʔa-momori ‘believe’
PPn *maqoli ‘true, real’ (POLLEX)
Pn Niuean mooli ‘true, sure’
Pn Anutan maori ‘indigenous, true, close of kin’ (EPn loan?)
Pn Emae māri ‘true, indeed, truth’
Pn Ifira-Mele māori ‘true, real’
Pn Pileni maoli ‘true; tell the truth’
Pn Rennellese māʔogi ‘right, true, real; exist’
Pn Tikopia maori ‘true, truth; feel sure of’
Pn West Futunan mari ‘true, truth, indeed’
PEPn *maoli ‘true, genuine; native, indigenous’
Pn Rapanui maʔori ‘skilled, old’
Pn Hawaiian maoli ‘true, real, native, indigenous’
Pn Marquesan maoʔi ‘indigenous’
Pn Tahitian māohi ‘native, indigenous’ (-h- unexpected)
Pn Tongarevan māori ‘local, aboriginal, traditional’
Pn Tuamotuan maori ‘indigenous’
Pn Rarotongan māori ‘of native origin, indigenous’
Pn Māori māori ‘indigenous, natural; mortal man as opposed to supernatural beings; fresh (of water)’
Pn Moriori mōri-ori ‘indigenous people of the Chatham Islands’
cf. also:
SES Gela mauri ‘living, real’
SES Lau mori ‘alive, real’
NCV South Efate mori ‘true’

POc *ma-qoni ‘true, real’
PT Balawaia moɣoni ‘true’
PPn *maqoni ‘true, real’ (POLLEX)
Pn Tongan moʔoni ‘true, genuine, real, intrinsic’
Pn Samoan moni ‘true, speak truth’
Pn Samoan (faʔa)maoni ‘true, faithful’
Pn Anutan mooni ‘true, as opposed to a lie’
Pn East Uvean moʔoni ‘true, certain’
Pn Sikaiana māoni ‘true, genuine’
Pn Takuu maoni ‘true, real’
Pn Tokelauan moni ‘true, sincere, honest’

PEMP *molaŋ ‘true, real, genuine’ has just one known non-Oceanic reflex, Buli molaŋ ‘correct, real, genuine, true’ (ACD).

PEMP *molaŋ ‘true, real, genuine’ (ACD)
POc *mola(ŋ) ‘true, real, genuine’
NNG Lukep mōl-mōl ‘true’
MM Nakanai imo-imola ‘talk that is true; the truth’ (i- unexplained)
SES Lau mola ‘true, real, abundant’
SES Arosi mora ‘original, true, real; customary’
SES Owa mʷora ‘true, real’
Mic Marshallese mʷōl ‘true’

The question mark against POc *moqi below refers to its form. If Takia mok is indeed a reflex, then medial *-q- should be reconstructed.

POc *moqi ‘true’
NNG Takia mok ‘true, real; very, truly’
NNG Dami mo-moi ‘true’
NNG Manam moi-moi ‘true’
PT Tawala moi- ‘true’
SES Arosi moi ‘true’7

A small number of forms meaning ‘true’, all Northwest or Southeast Solomonic, reflect a root *mana. It is tempting to associate these with POc *mʷane ‘straight, direct; flat, level’ (Vol. 2:213),8 and this is probably the origin of Gela mae-mane ‘correct’ below. However, neither forms nor meaning otherwise support this association. It is possible that these forms are cognate with PPn *mana ‘supernatural power’ (POLLEX) and reflect the term that Blust (ACD) reconstructs as POc *mana ‘power in natural phenomena; thunder, storm wind’. However, the glosses below suggest (i) that *mana/*ma-mana was a homophone of Blust’s reconstruction, and (ii) that the Simbo and Lau reflexes below reflect a conflation of Blust’s POc *mana ‘power…‘ and *mana ‘true’. Since all reflexes of the latter are located in the Solomons archipelago, it is difficult to know which interstage it should be attributed to.

MM Nehan mana ‘true’
MM Halia mana ‘true’
MM Teop mana ‘truth, meaning’
MM Teop (va)mana-mana ‘believe’
MM Simbo mana ‘true; powerful, potent, effective; gracious; to grant, be favourable; power’
SES Ghari mana ‘truth, true, correct’
SES To’aba’ita ma-mana ‘true, real’
SES To’aba’ita faʔa-mamane- ‘believe’
SES Lau ma-mana ‘efficacious; be true, come true, be fulfilled’
SES Gela mae-mane ‘correct’

The set below deserves mention because of its frequent occurrence in Table 25. It is restricted to SES languages, and there seems to be no consistent semantic difference between forms with and without *-ni.

PSES *utu, *utuni ‘true’
SES Bugotu [t]utuni ‘true’
SES Bugotu (va)utu-utuni ‘believe’
SES Gela utu ‘true’
SES Gela utuni ‘certainly, truly, right’
SES Gela (talu)utuni ‘believe’ (talu ‘put’)
SES Tolo utuni ‘true, correct’
SES Tolo (t)utuni ‘believe’
SES Longgu utuni ‘true’ (borrowing)
SES Longgu (naʔi)utuni ‘believe’ (naʔi ‘put’)
SES Arosi ū ‘true, real’
cf. also:
SES Longgu uðua ‘true’

Finally, the small set below has a curious distribution. Reflexes of PMP *bener occur in Western MP languages, but none are known in Oceanic outside Eastern Polynesian.

PMP *bener ‘true, righteous, honest’ (ACD)
POc *bono(r) ‘true, correct’
PPn *pono ‘true, correct’ (POLLEX)
Pn Māori pono ‘true; hospitable, bountiful; abundant; means, chattels, abundance’
Pn Māori ɸaka-pono ‘believe, admit as true’
Pn Marquesan pono ‘correct, proper, well done’
Pn Hawaiian pono ‘correct procedure, correctness’

5. Remembering⇫

Probably all Oceanic languages have terms for MEMORISE (‘commit s.t. to memory’) and for RECALL (‘remember s.t. /that…’), but these terms are usually complex lexemes, (§10.1). The glosses of their components are given henceforth in square brackets. Table 26 sets out terms for the two semantic frames in the four witness languages.

Table 26 Verbs of remembering in the four witness languages
MEMORISE RECALL
‘commit (s.t.) to memory’ ‘remember (s.t. /that …)’
Nakanai mata-toro [look-strong] hilo-tavu [see-towards]
To’aba’ita manata oli uri- [think back about]
Mwotlap dem sas [think find]
Wayan katoni- ‘put in box’, numi-lesu-ni- [think-back-TR],
bolani- ‘put in basket’ divi- ‘daydream, remember longingly’

The absence of MEMORISE lexemes in Mwotlap and To’aba’ita typifies their absence from many dictionaries. The data are so sparse that they will not be further considered here. The Wayan verbs are simple metaphors: katoni- is derived from kato ‘container with lid’ and bolani- from bola ‘coconut leaf basket, container with lid’.

The default POc RECALL verb was apparently *nonom, *nanam ‘think about s.t., remember s.t.’, reconstructed in §10.3. It encoded both RECALL and COGITATE frames. The only simple RECALL verb in Table 26 is the Wayan verb divi- ‘daydream, remember longingly’, but this includes the additional sense of longing, quite common in RECALL verbs in Oceanic languages.

The remaining RECALL terms in Table 26 are complex lexemes, and three of them begin with the language’s default COGITATE verb. In this they are typical of Oceanic RECALL terms outside Polynesia. It is possible that, for example, the ‘think + find’ sequence immediately below is of POc antiquity, but the data do not allow us to reconstruct the forms that occurred in this and other complex lexemes.

An effect of employing complex lexemes is that they may encode more specific meanings than English usually encodes with remember. Thus one sense of remember, as in ‘He managed to remember the address’, views remembering as finding a piece of information in one’s memory after a search, encoded by a SVC ‘think + find’:

PT Dobu nua loba [think find] ‘think and finally remember’
SES Kwaio manata dalia [think find] ‘remember, recall’
NCV Mota nom suar [think find] ‘think and find, recollect’
NCV Mwotlap dem sas [think find] ‘remember’
NCV Paamese mudem sāli [think find. out] ‘remember, discover’

These data imply the existence of a compound lexeme meaning ‘search one’s memory for s.t.’, and examples occur, but sometimes with rather vague glosses. Here and below, languages around the Vitiaz Strait replace ‘think’ or ‘mind’ with ‘eye’, giving a BPM.

NNG Bariai i-mata nanan [S:3SG-eye pursue] ‘remember’
NNG Kove mata-ɣu i-nana [eye-my S:3SG-pursue] ‘remember’
SES Kwaio manata fana [think hunt] ‘think about, remember’
SES Kwaio lada ʔōfia [dig. up look. for] ‘wander about, search for, try to remember’
NCV Mwotlap dem sɔsɔk [think look. for] ‘think hard in order to remember s.t.’
NCV Paamese mudem lēkati [think look. for] [VT] ‘try to remember’ (lē-kati [see-really] ‘look for’)

Remembering in the sense of casting one’s mind back, recalling and recollecting is often expressed by the sequence ‘think + go back’ or sometimes ‘think again’. Note below that Iduna has two syntactically different variants of the same expression. In one, nua- ‘mind’, a monovalent noun, is subject of the verb -nauye- ‘go back’. The other is a compound verb made up of the same morphemes.

NNG Mangap mata- i-miili [eye- S:3SG-go. back] ‘remember again’
PT Dobu nua-ila [mind go. back] ‘think of the past, reminisce’
PT Iduna nua- gi-nu-nauye- [mind- S:3SG-REDUP-go. back-] ‘remember, call to mind, think about’
PT Iduna -nua-nua-nauye- [think-think-go. back-] ‘remember, think about, consider, recall s.t.’
PT Tawala nugo-gae [think-go. up] ‘remember, recall’
MM Patpatar lik leh [think go. towards] ‘remember’
MM Tolai nuk-mule [mind again] ‘remember, recall to mind’
MM Nehan namana poluku [think again] ‘remember again, recall to mind’
MM Tinputz nat hah [know again] ‘remember’
SES Gela ganagana oli [think-go. back] ‘remember’
SES Tolo pada-visu- [think-go. back-] ‘remember’
SES To’aba’ita manata oli uri- [think go. back toward-] ‘think back to’
NCV Mota nom-kel [think back] ‘call to mind, remember’
NCV Mwotlap dēm lok [think again] ‘remember’
Fij Wayan numi-lesu-ni- [think-back-TR] ‘recall or think back on s.t.’

Remembering in the MEMORISE sense of holding something in one’s memory is expressed in a number of WOc languages by the sequence ‘think + hold’, or in Nehan by a simple ‘hold’ metaphor.

NNG Kove mata-xu vara [eye-my hold. tight] ‘I think of s.t., remember s.t’
PT Gumawana nuo-kavata [think hold] [VI] ‘remember’
PT Gumawana nuo-kavate [think hold] [VT] ‘remember s.t.. memorise s.t.’
PT Dobu nua-yai [think-hold. firmly] ‘remember’
PT Minaveha nua vi-avini [think S:3SG-hold] ‘remember s.o., s.t.’
MM Nehan saŋa dede [hold continually] ‘remember well’

Much the same concept is occasionally expressed by a ‘think + stay’ sequence:

PT Balawaia tuɣamaɣi-taɣo [think-sit.quietly] ‘remember, think of’
MM Patpatar lik kawase [think wait] ‘remember’
SES Lau manata tō [think stay] ‘remember’

In many Oceanic languages, serialisation and compounding have remained productive, and there are complex lexemes that appear to be quite localised:

NNG Mangap mata- i-ᵑgal [eye- S:3SG-pierce] ‘think of, remember’
NNG Tuam mata i-ᵑgal [eye S:3SG-pierce] ‘remember’
PT Gumawana nua-isi [think-break] ‘remember s.t.’
PT Iduna nua- -afole- [mind- -pierce] ‘remember, recall’
PT Iduna ua- -ʔakakili- [mind- -overbalance] ‘suddenly remember s.t.’
PT Tawala nugo-momota [think-pull. tight] ‘remember, hold in the heart’
MM Nakanai hilo-tavu [see-towards] ‘remember’
NCV Paamese mudem silati [think come. across. by. chance] ‘suddenly recall’

6. Forgetting⇫

Like terms for RECALL, many terms for forgetting are complex lexemes, the first component of which is either the default COGITATE verb or the body-part noun that the language uses for ‘mind’. The second component is a verb, the meanings of which are in several instances quite widespread. There are dozens of combinations in the data. A geographically well distributed combination is ‘think/mind’ + ‘lose’.

Adm Nyindrou bale- mani [neck lose] ‘forget, lose’
NNG Bariai mata- sapian [eye lose] ‘forget’
PT Balawaia tuɣa-rekwa [think-lose] [VT] ‘forget’
MM Patpatar lik luben se [think lose] ‘forget’
NCV Lonwolwol nɔ̄r helalɛ [think lose] ‘forget’

Another is ‘think/mind’ + ‘short’, where ‘short’ is apparently used metaphorically for ‘lacking’. The two terms below are from the opposite geographic extremes of MM.

NNG Mengen lau pogo [liver.my be.short] ‘forget’
MM Maringe ɣaðo kmoʔe [think be.short] ‘forget’

The existence of a verb meaning ‘not know’ in many Oceanic languages was noted in §10.2. It figures as the second component of the following lexemes.

NNG Takia ilo- -ŋaoŋ [inside- -not know] ‘forget’
PT Iamalele nua-fani [think-not.know] ‘forget’
PT Iamalele nua- -fani [mind- -not.know] ‘forget’
MM Maringe ɣaðo iho [think not.know] ‘forget’
SES To’aba’ita lio-dorā [look-not.know] ‘forget (about)’
SES Kwaio maa-bolosia [eye-not.know] ‘forget’

A number of complex lexemes glossed ‘forget’ have a verb meaning ‘leave, go away’ as one of their components, usually the second. However, some of these have glosses—‘abandon’, ‘leave behind’—that imply a conscious choice to forget.

Adm Baluan wot lilisek [go. away forget] ‘forget’
MM Nakanai tapa-taro [? -away] [VT] ‘forget, leave, behind, abandon’ (tapa apparently does not occur as a verb alone)
MM Maringe ɣaðo ɣosu [think leave. behind] ‘forget, leave behind; ignore; be unaware of’
NCV Mwotlap dem vɛtɛɣ [think leave] ‘forget, pardon, abandon, drop’
Fij Wayan numi-deini- [think-leave] ‘forget s.t., have s.t. slip one’s mind, be unable to remember s.t.’

Clark (2009:130) reconstructs a PNCV BPM *lolo- boŋi [mind night] ‘forget’, and infers that one component or the other has been replaced in various languages. He may well be right, but a more conservative inference is that a complex lexeme ‘mind’ + ‘night’ was present in early EOc. The terms for ‘night’ reflect either POc *rodrom ‘be dark, be night’ or POc *boŋi ‘night’ (vol.2:295–298). In some languages this BPM also has the sense ‘be ignorant’ (§11.3.4.1).

SES Sa’a maa rodo [eye night] ‘be blind, forget’
SES Ulawa sae rorodo [liver night] ‘forget’
NCV Mota lolo-pʷoŋ [inside-night] ‘ignorant, stupid, unenlightened; forget’
NCV Mwotlap lɔl-pʷoŋ [inside-night] [VT] ‘forget, ignorant’
NCV Nokuku lolo- ōra [inside- night] ‘forget, ignorant’
NCV Southeast Ambrym e- bovoŋ [? -night] ‘forget’9
NCV Port Sandwich na-lö- e-boŋ-boŋ-ini [ART-inside- it-REDUP-night-TR] ‘forget’
NCV Paamese ē- vo-boŋo [inside- night] ‘forget’
NCV Lewo sine- poni [guts- night] ‘forget’
NCV Lonwolwol lɔ- mʊ buŋ-buŋ [inside-? night] ‘forget’

PPn *nimo ‘vanish, forget’ perhaps reflects a metaphorical use of ‘vanish’ for ‘forget’.

PPn *nimo ‘vanish, forget’ (POLLEX)
Pn Tongan (ma)nimo ‘secret, underhand, surreptitious’
Pn Niuean nimo ‘forget’
Pn Niuean nimo(pō) ‘forget completely’ (pō ‘dark’)
Pn Samoan ni-nimo ‘completely forgotten’ (nimo ‘vanish, disappear’)
Pn Rennellese nimo ‘forget, vanish’

PPn *ŋalo uses the metaphor of a submerged (i. e. hidden) rock for ‘forgotten’.

POc *mʷaloq ‘submerged rock or coral reef, coral head’ (vol.2:108)
PPn *ŋalo ‘out of sight, disappeared, forgotten, lost’ (POLLEX)
Pn Tongan ŋalo [VSt] ‘be forgotten, sink, disappear from sight or memory’
Pn Niuean ŋalo ‘be lost, absent’ (faka-ŋalo-ŋalo ‘try to forget’)
Pn East Futunan ŋalo ‘forgotten’
Pn Samoan ŋalo [VSt] ‘forgotten’
Pn Tikopia ŋaro [VSt] ‘be lost (from sight or mind)’
Pn Tikopia ma-ŋaro-ŋaro ‘lost, gone out of sight’
Pn Māori ŋaro ‘disappeared, forgotten; be out of sight, invisible’
Pn Hawaiian nalo ‘disappeared, forgotten, lost’

7. Deciding⇫

The gloss ‘decide’ is rare in dictionaries of Oceanic languages, implying that deciding is not an Oceanic concept. One reason for this is that major decisions are traditionally made by consensus, for which—if one digs far enough—a term can be found. Its meaning, though, often includes the foregoing discussion as well as the decision.

NNG Takia awa- -tumani [mouth confer] ‘agree, decide together, come to consensus, take counsel (with each other)’
NNG Mapos Buang jō ɢaɢek [tie.knot speech] ‘decide, agree, to finish a discussion and come to a conclusion’
MM Nehan uel-halata [RECIP-discuss] ‘decide; discuss, decide together’
Fij Wayan boseti- ‘confer about s.t., meet to discuss or decide on s.t.’

No reconstruction can be made, and no consistent BPM pattern has been found.

On the rather rare occasions that one finds a term that appears to denote individual decision-making, it typically also includes either a reference to planning or to choosing. Indeed, the gloss ‘plan’ occurs rather more frequently than ‘decide’, but again no reconstruction is possible. ‘Choosing’, on the other hand, is clearly an Oceanic concept, and a verb can be reconstructed (§10.10).

When one searches a dictionary for ‘decide’, the gloss ‘undecided’ frequently turns up, and this is the topic of the next section.

8. Being undecided, of two minds⇫

The English idioms ‘be of two minds’ (this section) and ‘be of one mind’ (§10.9) translate as semantically similar BPMs in Oceanic languages.

Numerous expressions in Oceanic languages for ‘be undecided’ translate roughly as ‘be of two minds’. Expressions for ‘be undecided’ have been found in three of the four witness languages, and examples are given in Table 27.

The Nakanai example and the first To’aba’ita example are BPMs, with a body-part as subject and ‘two’ as predicate. The second To’aba’ita example is a compound verb, ‘mind’ + ‘two’, presumably derived from a BPM. The Wayan example also appears to be a BPM-derived compound, but here ‘two’ is replaced by ‘entangled (with weeds)’.

Table 27 Predicates of indecision in three witness languages

    • Nakanai (MM)
      ‘Thoughts are two.’
      la-gabutatala ilua
      ART-thoughts two
    • To’aba’ita (SES)
      ‘My mind is two.’
      manata-ku e=ruarua
      mind-my it=two
    • To’aba’ita (SES)
      ‘I (am) two minds.’
      nau ku=manata-ruarua,
      I I=mind-two
    • Wayan (Fij)
      ‘S/he (has) entangled thoughts.’
      Sā leŋaleŋā-rau
      s/he thinking-entangled

BPMs that are semantically similar to the top three examples in Table 27 are widespread in Oceanic languages (but seemingly infrequent in Vanuatu), and such a metaphor almost certainly occurred in POc.

NNG Lukep lo- ru [insides two] ‘be undecided’
NNG Takia ilo- ulalu [insides two] ‘doubt, uncertain, unsure’
NNG Yabem tɪtaʔ lulu [belly.his twofold] ‘be in doubt’
NNG Numbami tae-lualua [guts-two] ‘doubt, be of two minds’
NNG Buang kʷa lū [throats two] ‘undecided, doubtful’
NNG Buang ayo lū lū [feelings two two] ‘undecided, doubtful’
NNG Manam ilo- i-rua-rua [insides- 3SG-two-two] ‘doubtful, undecided, hesitating’
PT Dobu (e)nuana-lua [two-minds] ‘doubt’
PT Iamalele -nuana-luɣa [-mind-two] [VI] ‘undecided’
PT Iduna -nuanua-luɣa [-mind-two] [VI] ‘doubleminded, undecided’
PT Tawala nugo-lualuaga [mind-twofold] [VI] ‘confused, hesitant, undecided between two courses of action’
PT Misima nua-elelua [mind-twofold] ‘undecided’
PT Misima nua-lalabui [mind-twofold] ‘undecided, be of two minds’
MM Nakanai la-gabutatala ilua [ART-thoughts two] ‘of two minds, undecided’
MM Patpatar i-riruo lilik [be.two thought] ‘doubt’
MM Sursurunga ru i kən hol [two in her/his thought] ‘doubt’
SES Bugotu gāgana ruarua [thought two] ‘doubt, be undecided’
SES Sa’a sae rueruaʔa [liver twofold] ‘doubt’
SES Lau ro si lio [two of voice] ‘undecided, double-minded’
SES Lau manata rurua [mind two] ‘doubt’
SES ’Are’are manata-rua [mind-two] ‘divided in mind’
SES To’aba’ita manata- ruarua [mind- two] ‘be undecided, of two minds’
SES To’aba’ita manata-ruarua [mind-two] ‘be undecided, of two minds’
SES Arosi ahu-ruaruā [mind-two] ‘be in two minds’
SES Owa tako ruarua [mind two] ‘doubt’
NCV Mota nom-ruarua [mind-two] ‘be in two minds, hesitate, doubt’
Mic Kiribati nano-uoua [mind-two] [N] ‘doubt, perplexity’
Fij Bauan lomaloma-rua [insides-two] [N] ‘hesitation’
Fij Bauan lomaloma-rua-taka [insides-two-APPLICATIVE] ‘be in doubt about’
Pn Tongan loto-loto-ua [inside-inside-two] ‘of two minds, undecided’
Pn Samoan faʔa-lotu-lotu-lua [CAUS-inside-inside-two] ‘indecisive’
Pn East Uvean faka-loto-loto-lua [CAUS-inside-inside-two] ‘hesitant’
Pn Rarotongan ŋākau rua [guts two] ‘of two minds’

If by inference the POc body part in this context was *lalom ‘insides’ (§9.4), then the POc BPM *lalo- rua-rua is a plausible reconstruction.

In a few languages a reduplicated reflex of POc *rua ‘two’ with the sense ‘twofold’ serves alone as ‘of two minds’.

Mic Carolinian rɨarɨ [twofold] ‘be undecided, in doubt, of two minds’
Fij Rotuman ararua [twofold] ‘(habitually) indecisive’

Occasionally the BPM turns up with ‘many’ instead of ‘two’, indicating that in some languages at least the metaphor remained productive.

NNG Takia ilo- wei [insides- many] ‘be in doubt’
PT Dobu (e)nuana-yauna [mind many] ‘be undecided’
Mic Kiribati nano koraki [insides crowd] ‘indecision’
Mic Kiribati nano maiti [insides many] ‘perplexed’

9. Agreeing, being of one mind⇫

Metaphors for ‘agree, reach consensus, be unanimous’ fall into two patterns. The first roughly translates ‘be of one mind’. Its POc form may well have been parallel to that of POc *lalo- rua-rua ‘be of two minds’ (§10.8), but its reconstruction is obstructed by the fact that several POc forms for ‘one’ can be reconstructed (Lynch, Ross & Crowley 2002:72), and their distribution is not yet well enough understood to infer which form probably occurred in this BPM.

NNG Takia ilo- kisaek [insides one] ‘be of one mind, agree’
PT Iduna veʔa-nuwanuwa-saeʔya- [RECIP-mind-one-] ‘be of one mind with (s.o.)’
PT Kilivila nina-tala [mind-one] ‘be of one mind’
PT Motu lalo- tamona [insides one] ‘agree’
SES Bugotu lio- sikei [mind one] ‘of one mind, decided; resolute’
Mic Carolinian tipi-yew [one neck] [VI] ‘be of one mind, agree’
Pn Tongan loto-taha [insides-one] ‘unanimous, of one mind’
Pn Rennellese goto tasi [insides one] ‘agree’

Other languages employ a variety of complex lexemes meaning ‘mind together’ or ‘speak together’.

Adm Nyindrou sahou radra le [talk one only] ‘agreement, covenant’
NNG Takia awa- -tumani [mouth- -meet] ‘reach consensus, make collective decision’
PT Motu gwau-bou [speak-together] [VI] ‘agree’
SES Gela lio kolu [mind together] ‘agree’
SES Bugotu fari hagore [share speech] ‘agree’
SES Sa’a ruru wala [gather.together speech] ‘agree’
SES Owa tamasi faga-etagai [speak CAUS-one] ‘agree’
NCV Mota sara-tuwale [gather.together-one] ‘agree, meeting together’
Fij Bauan loma-vata [insides-together] ‘agree’
Fij Wayan lia vata [one together] ‘be unified, unanimous’

10. Choosing⇫

Choosing is a cognitive act, but it is one that has visible physical consequences, and it is perhaps for this reason that a POc etymon, *piliq (VI), *piliq-i- (VT) ‘choose, select, pick out’ , has enjoyed considerable continuity and relatively little replacement.

PAn *piliq ‘choose, select, pick out’ (ACD)
POc *piliq [VI] ‘choose, select, pick out’
POc *piliq-i- [VT] ‘choose, select, pick out’
NNG Takia -pili-an- ‘mark out, select, choose, pick up, deal out’
PT Gapapaiwa vine ‘choose’
PT Tawala win(agana) ‘choose, select’
PT Dawawa vine ‘choose’
PT Misima hili ‘choose (piece of material)’
PT Balawaia viriɣ-i ‘choose’
MM Sursurunga pilək ‘choose’
MM Ramoaaina pilak ‘choose, select’
SES Bugotu vili ‘choose’
SES Gela vili ‘choose, select; give a judgment’
SES Longgu vili- ‘choose, select, appoint’
SES Lau fili ‘choose, prefer’
SES Lau fili-s-ia ‘be chosen’
SES Sa’a hili ‘choose for one’s own, desire and take’
SES Sa’a hili-si ‘pick, choose’
SES ’Are’are hiri-si- ‘choose, pick out, select’
SES Arosi hiri ‘choose’
NCV Lonwolwol wɛl ‘choose’
PMic *fili ‘choose’ (Bender et al. 2003)
Mic Chuukese fiɾi- ‘choose, select; appoint’
Mic Puluwatese fili- ‘choose, select’
Mic Woleaian f-firi ‘choose, pick up, decide, select’
Mic Ponapean pil ‘choose, pick out, select’
Mic Mokilese pil ‘choose, select’
Fij Rotuman hili ‘pick out, choose, select’
Fij Wayan vili- ‘pick up (s.t.)’
Fij Bauan vili ‘pick up scattered things, as fallen leaves or fruits’
PPn *fili ‘choose’ (POLLEX)
Pn Tongan fili ‘choose, pick out, cull, select’
Pn Niuean fi-fili ‘choose, select’
Pn East Futunan fili ‘choose’
Pn Samoan fili ‘choose’
Pn Kapingamarangi hili ‘choose, select; choice’
Pn Nukuoro hili ‘pick from among several, choose; be choosy’
Pn Rennellese higi ‘choose, select; be choosy, selective’
Pn Rarotongan iri ‘select, choose, pick out, name’
Pn Māori ɸiri ‘select, choose’

11. Learning and teaching⇫

To teach someone something is to cause them to learn it, and the learner then either knows what has been taught or knows how to do something. This causative relationship is explicit in many Oceanic languages. Occasionally it is expressed simply by using the same transitive verb for learning something and teaching something, as in the instances below:

SES Tolo sasani- ‘learn, educate, instruct’ (cf. sasani (VI) ‘learn, go to school’)
SES To’aba’ita toʔo- ‘learn s.t.; teach s.o. s.t.’
Fij Wayan vuli-ði- ‘study, learn s.t., teach s.o.’ (vuli (VI) ‘study, learn’)
Pn Samoan aʔo ‘learn, teach, train’

More often (Table 28) it is expressed by attaching a causative prefix, usually a reflex of POc *pa[ka]- CAUSATIVE, to a root meaning ‘know’ (§10.2) or ‘learn’. Since to learn is often synonymous with ‘come to know’, the English distinction between ‘know’ and ‘learn’ is not relevant in this context.

The Bariai (NNG), Misima (PT) and Halia (MM) items above, along with Dawawa (PT) wai-kata-i ‘show’, suggest that there was a PWOc causative *paka-qataq-i- ‘teach, cause to know, show’, formed from *qataq-i- ‘know, understand, realise (that)’ (§10.2).

Another semantic dimension of verbs of learning and teaching arises out of teaching styles in traditional Oceanic communities. A young person learned how to do something by watching an older person and imitating them, and this is reflected in the fact that POc *towa ‘imitate, learn by imitation’ and POc *usuri/*usawiri ‘imitate’ below both have ‘imitate’ as their primary sense. The causatives formed from them, however, mean ‘teach, instruct’, presumably by demonstration.

Lou, Baluan and Manam reflexes of intransitive POc *towa ‘imitate, learn by imitation’ reflect a transitive *towa-(a)kini- ‘learn (s.t.) by imitation’ formed with the suffix *-(a)kini (§1.3.5.2). Curiously, PT reflexes of *towa are formed with a causative prefix, but mean ‘imitate’ rather than ‘teach’.

Table 28 Verbs of teaching formed with a causative prefix
NNG: Bariai pa-oatai ‘teach’ oatai ‘know, possess knowledge’
NNG: Bariai pa-nanale ‘teach’ nanale ‘learn; be accustomed to, get used to’
NNG: Mangap -pa-kilaala ‘teach, help to understand’ kilaala ‘know well, be aware, understand’
NNG: Mangap pa-ute ‘teach’ -ute ‘know, know how to’
PT: Dobu e-ʔita ‘teach, show, train’ ʔita ‘see, look’
PT: Misima a-atena ‘teach’ ate(na) ‘know, understand’
PT: Balawaia vaɣa-riba ‘teach, inform’ riba ‘know’
PT: Motu ha-diba-ia ‘teach, learn’ diba_ ‘know, understand’
MM: Ramoaaina wer ‘teach’ wa-wer ‘learn; teach’
MM: Halia h-atatei ‘learn, begin to know’ atei ‘know’
SES: ’Are’are haʔa-usuri- ‘teach, instruct’ usuri- ‘follow, copy, imitate’
SES: Kwaio faʔamanatā ‘teach, advise’ manata ‘think, reason, know’
SES: Owa faga-usuri ‘teach s.o.’ usuri- ‘imitate s.o.’
Mic: Carolinian a-xulē-y ’teach (s.o.) xule ‘learn, know’
Fij: Bauan vaka-vuli-ði- ‘teach a person s.t., make s.o. learn s.t.’ vuli-ði- ‘study, learn s.t., teach s.o.’
POc *towa [VI] ‘imitate, learn by imitation’
POc *towa-(a)kini- [VT] ‘learn (s.t.) by imitation’
Adm Lou to-ek ‘show’
Adm Lou to-to-ek ‘show how’
Adm Baluan tou-ek ‘show; teach’
NNG Manam to ‘learn’
NNG Manam to-aka ‘imitate, copy, mimic, mock’
NNG Sio towo ‘demonstrate; show how’
PT Gumawana va-to-towa-na ‘imitate’
PT Bunama he-to-towa-ne ‘copy, imitate’
PT Dobu e-to-towa-na ‘copy, mimic, imitate’
MM Bola tovo ‘learn’
NCV Paamese te-toho-ni ‘imitate, copy’
NCV Lewo tou-towo [VI] ‘measure, imitate’
NCV Lewo tou-tou-ni [VT] ‘measure, imitate’

At first sight, the set below appears to reflect the POc root *usuri, but the Ramoaaina and Nehan reflexes suggest that formal reconstruction is more complicated. The Nehan root sairi contains no internal -u-. The fact that NW Solomonic languages lose -w-. but not -u-, suggests that sairi reflects usawiri (with unpredicted loss of initial u-). Ramoaaina loses -s-, and its expected reflex of *usuri would be †ūr, rather than actual wer, which also attests to the presence of *-w- followed by an unrounded vowel.

POc *usuri, *usawiri ‘imitate’
POc *pa[ka]-usuri, *pa[ka]-usawiri ‘teach, pass on’
MM Patpatar ha-usur [VT] ‘teach’
MM Patpatar hara-usur [VI] ‘learn’
MM Ramoaaina wer ‘teach’
MM Ramoaaina wa-wer ‘learn; teach’
MM Nehan ua-sairi ‘copy’
MM Nehan uala-siri ‘teach; understand, train, skill’
SES Longgu (gere)usuli- ‘copy writing’
SES Lau usuli- ‘copy; take after, resemble’
SES ’Are’are usuri- [VT] ‘follow, copy, imitate’
SES ’Are’are haʔa-usuri- ‘teach, instruct’
SES Arosi usuri [VT] ‘hand on a tale’
SES Arosi haʔa-usuri ‘teach, instruct; teacher’
SES Owa usuri- [VT] ‘imitate s.o.’
SES Owa faga-usuri [VT] ‘teach s.o.’
NCV Mota usur ‘pass on, relate’
cf. also:
MM Siar ariri ‘learn’ (first -r- for †-s-)

Reflexes of another term for ‘learn’ are known only from Gapapaiwa and from Polynesian languages.

POc *akop ‘learn’
PT Gapapaiwa akova ‘learn, know, understand’
PPn *ako ‘acquire mentally, learn, teach’ (POLLEX)
Pn Tongan ako [VI] ‘learn, study; teach, train in’
Pn Tongan ako-naki [VI] ‘teach, give instruction’
Pn Niuean ako [VI] ‘learn’
Pn Niuean faka-ako [VT] ‘teach, learn, teach yourself’
Pn Samoan aʔo ‘learn, teach, train’
Pn Tikopia ako ‘learn’
Pn West Futunan ako ‘learn, try, attempt’
Pn Mangarevan ako ‘prove, try, exercise, practise’
Pn Hawaiian aʔo ‘learn, teach’
Pn Māori ako ‘learn, teach’

The primary meaning of POc *[ña]ñau appears to have been ‘teach’, perhaps centring on parents or seniors instructing children orally about their responsibilities (cf Lukep, Sursurunga and Kwaio glosses) and/or showing them how to perform traditional practices (cf Gela and Lau glosses).

The form of *[ña]ñau is open to question. If the Nyindrou term is indeed a reflex, then the POc consonant was *ñ. But if the Nyindrou term is not a reflex, and the Titan term listed under ‘cf. also’ is a proper reflex, then the form was *[na]nau.

POc *[ña]ñau ‘teach, learn’
POc *pa[ka]-[ña]ñau ‘teach’
Adm Nyindrou ñowoña ‘reveal, point out’
NNG Kairiru -nanou-i ‘teach’
NNG Lukep -nōnō ‘teach about traditional responsibilities’
NNG Sio (pa)nana ‘teach’
MM Sursurunga (i)nau ‘instruct, charge (as parent to child)’
SES Gela naunau ‘teach a craft, teach a dance; try, practise; imitate’
SES Longgu nau-a ‘show s.o., teach s.o.’
SES Lau (fā)nanau ‘train, teach by practice’ (†nanau not recorded)
SES ’Are’are nao ‘turn, point, aim towards’
SES ’Are’are naohi- ‘point at, to aim at’
SES Kwaio nanau ‘learn about, learn’
SES Kwaio (faʔa)nanau-a ‘teach, lecture’
SES Sa’a (sae)nanau ‘be taught, be wise’ (sae ‘liver’)
NCV Mota (vata)nau ‘learn, teach, by practice’
cf. also:
Adm Titan ananowe, anano-ani ‘show, teach’

12. Conclusion⇫

For many cognition frames no POc term can be reconstructed. Apparently because their meanings are abstract, their lexical replacement rate is considerably higher than for items with less abstract meanings (§9.6). As we have noted, abstract states and activities tend to be encoded metaphorically as complex lexemes.

At the same time, it is reasonable to infer that, for example, the ‘think + find’ SVC pattern for ‘remember’ in §10.5 is quite probably of POc antiquity, as it occurs in widely distributed languages. However, the data do not allow us to reconstruct the forms that occurred in this and other complex lexemes, and so the possibility of independent parallel innovation cannot be excluded. Thus for remembering (§10.5), forgetting (§10.6) and being of one mind (§10.9) no forms are reconstructed, but complex lexemes are described, as they give us some insight into how POc speakers conceived these cognitive activities. For deciding (§10.7) not even a consistent pattern of complex lexemes is found, and the same is true of hoping and expecting, which are omitted here.

Notes⇫