Information and examples are from Seifart (2011; 2012). The following list is extracted from Seifart’s (2011:66–86) identification of Bora forms in Allin’s (1976:382–458) list of about 1,590 Resígaro words, selecting those affixes that are attested with at least one native Arawakan stem. Added to this selection of forms is the augmentative marker and the dative case marker, which do not occur in the word list, but are widely attested in other parts of Allin’s (1976) work as well as in Seifart’s (2009; 2012) data. Note that Seifart (2012) provides a different, although largely overlapping, list of borrowed affixes, namely those that are attested in a given corpus of Resígaro text (Seifart 2009).

40 classifiers and gender markers (38 classifiers predominantly used with inanimates, 2 feminine gender markers) (one native form in the same slot: ‑gi ‘non‑feminine’)

-ba ‘classifier.thing’, e.g. aʔmitʰoótsibá ‘belly’

-baba ‘classifier.bag’, e.g. hiiʔpabábɯ́ ‘sock’

-dʒihɯ ‘classifier.powder’, e.g. hípodʒíhɯ́ ‘powder’

-dʒiiʔo ‘classifier.broom’, e.g. híβeʔdʒiiʔó ‘(long) hair’

-gaaʔdʒó ‘classifier.raft’, e.g. aβáanagaaʔdʒó ‘raft’

-gahɯ ‘classifier.ridge’, e.g. ín̥igahɯ́ ‘eyebrow’

-gɯ ‘classifier.plank’, e.g. boeʔkʰóótsígɯ́ ‘paddle’

-ha ‘classifier.cover’, e.g. aʔmitʰoótsihɯ́ ‘kitchen’

-hi ‘classifier.disc’, e.g. epíitsíhí ‘axe’

-hɯ ‘classifier.tube’, e.g. okóniigihɯ́ ‘rifle’

-i ‘classifier.stick’, e.g. aβáanaí ‘stick’

-iko ‘classifier.nest’, e.g. nanáanáʔiikó ‘shoot, new growth (of a pineapple)’

-iʔikɯ ‘classifier.framework’, e.g. aápiíʔíkɯ́ ‘skeleton’

-iʔo ‘classifier.point’, e.g. henákoííʔo ‘horn’

-kaahɯ ‘classifier.swamp’, e.g. iteeβikaahɯ́ ‘place where aguaje palms grow’

-ko ‘classifier.pointed’, e.g. dʲɯííʃikó ‘hausai palm’

-kooʔa ‘classifier.shaft’, e.g. botoʔkʰoókooʔɯ́ ‘broom’

-kɯba ‘classifier.leg’, e.g. hiiʔpákɯba ‘leg’

-mi ‘classifier.transport’, e.g. okóniigimí ‘launch, boat’

-miiʔo ‘classifier.shell’, e.g. an̥oógí eémámííʔo ‘tapir skin’

-paahi ‘classifier.hollow’, e.g. híβeʔpaahí ‘skull’

-pahtsi ‘classifier.hole’, e.g. henákopásí ‘earring’

-pako ‘classifier.water’, e.g. keʔβigipákó ‘brandy’

-peko ‘classifier.night’, e.g. nokótsapékó ‘the night before last’

-ɯ ‘classifier.round’, e.g. hímiɯ́ ‘seed’

-ɯɯʔo ‘classifier.string’, e.g. epíipíɯɯʔó ‘liana cord’

-ʔaakɯ ‘classifier.pillar’, e.g. hiímiaakɯ́ ‘pillar in house construction’

-ʔaami ‘classifier.leaf’, e.g. apánáʔaamí ‘leaf’

-ʔasi ‘classifier.patio’, e.g. híβeʔásí ‘crown (of the head)’

-ʔe ‘classifier.trunk’, e.g. áɲaahíʔé ‘cashew tree’

-ʔehɯ ‘classifier.hole’, e.g. hooniʔéhɯ́ ‘well (waterhole)’

-ʔi ‘classifier.bunch/classifier.river’, e.g. pipíipíʔí ‘bunch of Guilelma (palm) fruits’

-ʔidʒo ‘classifier.pot’, e.g. itsaaʔniʔídʒó ‘earthware pot, pitcher, jug’

-ʔo ‘classifier.oblong’, e.g. ʃakooʔgíʔó ‘banana fruit’

-ʔoohɯ ‘classifier.cylinder’, e.g. hipon̥oótsíʔoohɯ́ ‘prison’

-ʔooβɯ ‘classifier.chunk’, e.g. keétséʔooβɯ́ ‘lantern, flashlight (torch)’

-ʔosi ‘classifier.hand’, e.g. apáʔmíʔosí ‘left hand’

-βɯɯ́dɯ ‘classifier.chunk’, e.g. aβáanaβɯɯdɯ́ ‘log’

-dʒe ‘feminine singular’, e.g. -βanaaʔdʒé ‘sister-in-law (sister of sister-in-law or brother-in-law)’

-pidʒe ‘feminine singular’, e.g. pʰaipídʒé ‘old woman’

6 number markers (one native form in the same slot: ‑ne ‘plural for humans’)

‑mɯ ‘plural for animates’, e.g. anadómɯ ‘macaws’

‑ʔa ‘plural for animates in certain pronominal expressions’, e.g. faʔá ‘we (inclusive)’

‑hi ‘plural for inanimates’, e.g. tsikíʔhɯ́hí ‘silly thing (lit. ‘empty words’)’

‑kɯ ‘dual for inanimates’, e.g. hamáakábaakɯ́ ‘hammocks (dl.)’

‑mɯpi ‘dual feminine’, e.g. pʰaimɯpi ‘two old women’

‑mɯsi ‘dual animate’, e.g. anáadómɯsi ‘two macaws’

4 forms that are not interrelated with other borrowed affixes

‑a‑ ‘belonging to’, e.g. ginomomináagi ‘white person (lit. one belonging to the white people)’

‑kobɯ ‘augmentative’, e.g. keéʔʃékobɯ ‘big cow’

‑ke ‘dative case’, e.g. náagi‑ké ‘[she gave a fruit] to her brother’

‑ʔi ‘verb marker (marking the main verb when in combination with the auxiliary verb kʰɯ́ ‘do’)’, e.g. kaaʃodʒáʔi kʰɯ́ ‘to like, to want’

Information and examples are from McConvell and Meakins (2005) and Meakins (2007; 2011). McConvell and Meakins (2005) argue that Gurindji Kriol emerged from code‑switching between Gurindji and Kriol, in which English‑based Kriol was the matrix language, and that therefore elements from Gurindji can be considered as borrowings. Only forms for which there are sufficient indications that they are used with native, English‑based stems, are counted here, e.g. the borrowed plural marker ‑rrat is not counted here because it “is only found on demonstratives of Gurindji origin” (Meakins 2007:386).

2 markers for core cases (Meakins 2007:367)

‑ngku ~‑tu ‘ergative’, e.g. diya‑ngku ‘the deer (threw)’, boi‑ngku ‘the boy (was minding)’, gel‑tu ‘the girl (gave)’

‑yu ~ ‑wu ~ ‑u ‘dative’, e.g. Malingu‑wu ‘to Malingu’; an additional example, jikinfaul‑u ‘for the chicken’, is a less clear because it dates from the 1970s, when loans were not established yet.

3 markers for peripheral cases

‑ngka ~ ‑ta ‘locative’, diya‑ngka ‘on the deer’, tri‑ngka ‘on the tree’

‑ngkirri ~ ‑jirri ‘allative’, e.g. motika‑ngkirri ‘to the car’

‑nginyi ‘ablative’, e.g. klif‑nginyi‑ma (cliff‑ablative‑topic) ‘off the cliff’, hol‑nginyi‑ma (hole‑ablative‑topic) ‘from the hole’, juwingjuwing‑nginyi ‘from the swing’

3 number markers

‑kujarra ‘dual’ (no examples of hybrids)

‑walija ‘paucal’, e.g. hook‑walija ‘a few hooks’, bi‑walija ‘a few bees’

‑purrupurru, ‑nyarrara, ‑nganyjuk ‘associative plural’ (no examples of hybrids)

2 nominalizers

‑ny ‘nominalizer’ (no examples of hybrids)

‑kaji ‘agentive noun’, e.g. toktok‑kaji (talk‑agentive) ‘recorder’, mok‑kaji ‘cigarette lighter’

4 nominal suffixes affecting information structure

‑rni ‘only’, e.g. leg‑ta‑rni (leg‑locative‑only) ‘right through his leg’

‑ma ‘topic’, e.g. klif‑nginyi‑ma (cliff‑ablative‑topic) ‘off the cliff’, mami‑ngku‑ma (mother‑ergative‑topic) ‘the mother (got)’

‑rla ‘focus’, e.g. hiya‑rla ‘here’

‑rayinyj ‘alone’ (no examples of hybrids, not treated by Meakins as belonging to this set of forms)

2 proprietive/privative suffixes

‑yawung ‘proprietive’, e.g. motika‑yawung ‘their car’, femli‑yawung ‘his family’

‑murlung ‘privative adjectivizer’, e.g. kuloj‑murlung ‘naked’

2 non-interrelated suffixes

‑kari ‘another’, e.g. mor‑kari ‘more’, najan‑kari ‘another’

‑k, ‑pijik ‘inchoative, attaching to nouns and verbs’ (no examples of hybrids)

Information and examples are from Golovko and Vakhtin (1990) and Sekerina (1994:22–29), the most complete descriptions and the only ones available in English. Some additional, and partially alternative, description is given by Thomason (1997). The situation may be summarized as follows: “the derivational and nominal inflection system of CIA [i.e. Copper Island Aleut, FS] is entirely of Aleut origin, but the verbal inflectional system is that of Russian” (Sekerina 1994:22). A total of 15 borrowed affixes are productively used on native stems.

6 present tense subject suffixes (Golovko and Vakhtin 1990:98, 104, 107–112; Sekerina 1994:22–29)

‑yu ‘first singular’, e.g. qɑ̄=yu ‘I eat’, sū=yu ‘I take’

‑iŝ ‘second singular’, e.g. qɑ̄=iŝ ‘you eat’, sū=iŝ ‘you take’

‑it ‘third singular’, e.g. qɑ̄=it ‘he/she/it eats’, sū=it ‘he/she/it takes’, cali=it ‘he fishes’

‑im ‘first plural’, e.g. qɑ̄=im ‘we eat’, sū=im ‘we take’

‑iti ‘second plural’, e.g. qɑ̄=iti ‘you (pl) eat’, sū=iti ‘you (pl) take’

‑yut ~ ‑yat ‘third plural’, e.g. qɑ̄=yut ‘they eat’, sū=yut ‘they take’

6 past tense inflectional suffixes (Sekerina 1994:22–29; Golovko and Vakhtin 1990:108)

‑l ‘past tense’, e.g. ukuxtɑ̄=l ‘he saw’, ayxacɑ̄=l=a=ya ‘I started’, ayxɑ̄=l=ya ‘I travelled’, ukuxtɑ̄=l=i ‘they saw’, sū=l ‘he took’, sū=l=i ‘they took’

‑i ‘past plural’, e.g. anĝaĝí=l=i=mi ‘we lived’, ú=l=i=mi ‘we were’, ukuxtɑ̄=l=i ‘they saw’, sū=l ‘he took’, sū=l=i ‘they took’

‑ya ‘first singular’, e.g. ayxacɑ̄=l=a=ya ‘I started’, ayxɑ̄=l=ya ‘I travelled’, ukuxtɑ̄=l ‘he saw’

‑ti ‘second singular’, e.g. mɑ̄=l=ti ‘you did’,

‑mi ‘first plural’, e.g. anĝaĝí=l=i=mi ‘we lived’, ú=l=i=mi ‘we were’

‑vi ‘second plural’ (no examples of hybrids)

(third person is unmarked)

1 optional subject feminine gender suffix (Golovko and Vakhtin 1990:108–109)

‑a, used after the borrowed past tense marker ‑l, e.g. ayxacɑ̄=l=a=ya ‘I started’

1 infinitive marker

‑t’ ‘infinitive’, suffixed to main verbs for plural formation which combine with a borrowed auxiliary, e.g. taana ni‑buud‑ish ukuu‑t’ ‘You won’t see the land’ (Sekerina 1994:25). This marker may be considered part of finite verbal inflection and thus as being morphosyntactically related to the other morphology.

1 negative verbal prefix (Golovko and Vakhtin 1990:102, 104, 109)

ni‑ ‘negative’, e.g. ni=túta=qaĝī=l ‘he would not listen’, ni=sūy ‘don’t take!’

Information and examples are from Pakendorf (2010). Pakendorf (2010) analyzes data from Kałużyński (1962), Korkina et al. (1982) and from her own corpus and elicited data. She concludes that there are 14 affixes borrowed from Mongolic languages that are “currently still relatively productive” and used on native Turkic stems. A few other borrowed affixes are attested, but these are “previously productive, or marginally productive”. The following 14 are the ones that Pakendorf (2010) characterizes as “currently still relatively productive” (see also Pakendorf (2012). (Note, however, that Pakendorf counts more borrowed affixes because she accepts Korkina et al.’s (1982) claim that other affixes are productive, while these are not attested as hybrid formations).

5 deverbal nominalizers

‑AːččI (also functions as habitual marker), e.g. tut‑aːččï ‘(he) build (it)’

‑AːhIn, e.g. ottoː‑hun ‘the hay‑making’

‑BIl, e.g. öröbül ‘Sunday’ (from öröː‑ ‘to rest a day’)

‑lAŋ, e.g. kisteleŋ ‘secret’ (from kisteː‑ ‘hide’)

-ltA, e.g. terilte ‘organization’ (from terij‑ ‘equip, organize’)

4 deverbal adjectivizers

‑GAj, e.g. bïtarχaj ‘small’ (from bïtarïj‑ ‘crumble, smash, pulverize’)

‑ɣAr/‑gIr, e.g. laspaɣar ‘broad’ (from laspaj‑ ‘be/seem too broad and fleshy’)

‑(I)mtAɣaj, no attestations in Pakendorf’s corpus, but ‘productive’ according to (Kałużyński 1962:92; Korkina et al. 1982:165)

‑mtIA, no attestations in Pakendorf’s corpus, but nevertheless “mainly found with Turkic stems” (Pakendorf 2010), and “relatively productive” according to Korkina et al. (1982:165)

5 non-interrelated suffixes

‑rɣAː ‘verbalizer’, e.g. küːhürgeː ‘consider oneself strong’ (from küːs ‘strength’)

‑TA ‘multiplicative derivative of numerals’, e.g. biːr‑diː‑te ‘once’

‑(I)ččI ‘deverbal adverbializer’ (source form is a Mongolic imperfective converb), e.g. ergičči ‘roundabout’ (from ergij‑ ‘return, turn around’)

‑t(tAr) ‘plural’, e.g. ïaraχa‑ttar ‘the heavy ones’

‑Aːt: ‘immediate precedence converb’, e.g. orgut‑aːt ‘as soon as it boiled’

Information and examples are from Boretzky (1994) and Boretzky and Igla (1991). Kalderash Romani has also borrowed morphemes from Greek and from various Slavic languages, but in each case fewer than from Romanian.

1 adjectivizer

‑icios, e.g. kalits̆oso ‘blackish’

1 diminutive

‑uljeco (and many phonologically related allomorphs), e.g. foruljeco ‘little town’ (example from Kalderash spoken in Serbia)

6 agent nouns/professions, including feminine

‑aš(i) (could be from Slavic languages), e.g. tsir‑aš ‘circus man’

‑tori, e.g. diilibatori ‘singer’

‑anka, e.g. amerikánka ‘American woman’ (example from Kalderash of France)

‑àsa, e.g. amperacása ‘empress’ (example from Kalderash of France, amperac‑ could also be Romanian)

‑ojka (no examples of hybrids)

‑twára, e.g. farmec̆twára ‘witch’ (example from Kalderash of France)

1 ordinal number formation

‑leä, e.g. dúitoleä ‘for the second time’ (example from Kalderash of Sweden)

1 indefinite pronoun formation prefix

vare‑, e.g. vareko ‘anybody’ (example from Kalderash spoken in Serbia)

1 privative verbal derivational prefix

des‑, e.g. dezmekljol ‘melt’ (example from Kalderash spoken in Serbia)

Information and examples are from Pakendorf (2012), see also Pakendorf (2009; 2010), who counts fewer borrowed affixes because of adopting different criteria. The ones given here are all attested with native stems, even though examples of hybrid formations are only given for a few here. Note that there is some formal overlap in the person-number markers of the three different paradigms.

4 mood markers

‑jAktAːk ‘necessitative’

‑dAg ~ ‑dAk ‘assertive’

‑A ~ ‑Ar ‘indicative present tense’, e.g. ọrọn em‑e‑j‑er (reindeer/come‑epenthetic‑connective‑present.participle) ‘the reindeer are coming’

‑jAk ~ ‑jAːg ‘hypothetical mood’

5 subject markers for necessitative mood and indicative mood (out of 6 such subject markers), combining with ‑jAktAːk ‘necessitative’ and ‑A ~ ‑Ar ‘indicative present tense’. Third person singular is unmarked.

‑pIn ~ ‑bIn ‘first singular’

‑kIn ~ ‑gIn ‘second singular’, e.g. em‑e‑j‑egin (come‑epenthetic‑connective‑present.2sg) ‘you come’, ọrọn em‑e‑j‑er (reindeer/come‑epenthetic‑connective‑present.participle) ‘the reindeer are coming’

‑pIt ~ ‑bIt ‘first plural’

‑kIt ~ ‑kI‑hnAn ‘second plural’, e.g. huː ọlọk ụtal‑a‑jaktaːkkịt (2pl/emphasis/understand‑epenthetic‑necessitative.2pl) ‘you have to understand’. (Note that the form ‑hnAn used by one speaker on various occasions is native Ėven and that the borrowed element ‑kIt is truncated in this form.)

‑llAr ‘third plural’ (Note that in the necessitative mood, native Ėven ‑A‑l ‘third plural’ is used.)

5 subject markers for assertive mood (all 5 borrowed), combining with ‑dAg ~ ‑dAk ‘assertive’. No second person plural form occurs in the corpus, since the assertive mood gives a strong nuance of emphasis to the assertion that would not be used with reference to the addressee.

‑Im ‘first singular’, e.g. haː‑j‑bat bọl‑lagịm (know‑connective‑negative/auxiliary‑assertive.1sg) ‘I did not know’

‑Iŋ ‘second singular’, e.g. neː‑j‑er bọl‑lagịŋ (put‑connective‑present.participle/auxiliary‑assertive.2sg) ‘you put (it)’

‑A ‘third singular’

‑pIt ‘first plural’

‑tArA ‘third plural’

5 subject markers for hypothetical mood (all 5 borrowed), combining with ‑jAk ~ ‑jAːg ‘hypothetical mood’. First and second person plural forms do not occur in the corpus.

‑pIn ‘first singular’

‑kIn ‘second singular’

‑In ~ ‑n (with deletion of final ‑k of ‑jAk) ‘third singular’, e.g. em-u-jen (come-causative-hypothetical.3sg) ‘(he) might bring’

‑tArIn ‘third plural’

‑kA ‘impersonal’

1 non‑interrelated suffix

‑Is ‘ordinal numeral formation marker’, e.g. dig‑is ‘forth’

Note that two further Sakha affixes are found in the corpus, but only rarely occurring on Ėven stems, or in contexts where Sakha might be the matrix language, in which case Ėven words receive Sakha inflection. Therefore, they are not counted here. These are (i) -An ‘perfective (sequential) converb’, e.g. ič-e-j-en (see-epenthetic-connective-perfective.converb) ‘yes, they do see’; and (ii) -BAt ‘negative present’, which stands in paradigmatic opposition to -Ar ‘indicative present tense’, e.g. haː-j-bat (know-connective-negative) ‘(I) did not know’.

Information and examples are from Doerfer (1967), who mainly analyzes data from Rastorgueva (1964), see also Boretzky (2004:1651). Information on the etymology of stems was provided by Don Stilo (personal communication 2013).

2 nominalizers

‑či ‘agent noun’, e.g. χizmat‑či ‘servant’, “productive” according to Doerfer (1967:54)

‑lik ‘abstract noun’, e.g. χoǧa‑lik ‘farm, estate’

5 case suffixes (out of 6 Northern Tajik case suffixes). Some of the borrowed case markers seem to be used alternatively with native case markers, the genitive/accusative case is the only case for which there is only a native, and no borrowed case marker (Doerfer 1967:62).

‑ga ‘dative’, e.g. χåna‑ga ‘to the house’, used alternatively to χåna‑va, with a native Persian case marker. ‑ga is maybe only used with Persian verbs and nominalized verbs, but there is a clear tendency that it is on the rise (Doerfer 1967:62).

‑dan ‘ablative’, e.g. yakom klasašdan (first/class-3sg.possessor-ablative) ‘from the first class/grade’ (Doerfer 1967:54)

‑nda ~ ‑da ‘locative’, e.g. inǧanda ‘here’ (Doerfer 1967:54)

‑gača ‘terminative’, e.g. tå bist‑u‑haftom sål‑gača ‘up to the 26th year’ (Doerfer 1967:54)

‑dak ‘equative/comparative’ (Doerfer 1967:16, 56), no examples given, but described as part of a borrowed paradigm of case markers (Doerfer 1967:62)

2 derivational suffixes combining with numerals

‑gina ‘quantitative limiter’, e.g. yak som‑gina ‘one Ruble more [e.g. more expensive, cheaper]’

‑la, ‑lamiš ‘collective numeral formation’, e.g. duttalamå ‘both of us’, čårtalamiš ‘(with) all four’

2 suffixes not related to other borrowed morphology

‑ča ‘adverbializer’, e.g. urus‑ča ‘in Russian’

‑råq ‘comparative’, used together with the native Tajik comparative suffix, e.g. tez‑tar‑råq or tez‑råq‑tar ‘faster’

Doerfer (1967:18, 19, 25, 39, 56) mentions four other cases of morphological borrowing. They are not counted here for the following reasons: First, two of these appear to be free forms, although maybe derived from Turkic affixes: ikin ‘dubitative’ and či ‘emotional particle’. Second, the remaining two markers are explicitly described as combining only with Turkic stems: inči ‘ordinal number formation’ and miš ‘verbal noun formation’.

Information and examples are from Jones (1984), Verhaar (1984), Tadmor (2009), and personal communication with Uri Tadmor, David Gil, John Bowden, and a number of Indonesians working at the Jakarta Field Station of the Max Planck Institute for Evolutionary Anthropology in 2011, additionally supplemented with a number of hybrid formations and etymological details from Wiktionary (https://en.wiktionary.org/, accessed: 18.06.2024]). Jones (1984:7) states that “some half dozen affixes derived from Sanskrit are productive; most have come into vogue within the last few decades”. All of these are also mentioned by Verhaar (1984:20–21), who gives many more. Below, only affixes that are attested in hybrid formations, i.e. with non‑Sanskrit stems, are given.

16 miscellaneous nominal prefixes with non‑Sanskrit stems are attested (most of them native Indonesian, some from Javanese)

adi‑ ‘super’, according to Verhaar (1984:20–21) an affix of Sanskrit origin, compatible with native Indonesian stems, e.g. adibintang (bintang ‘star’) ‘superstar’, adipenghantar (penghantar ‘conductor’) ‘superconductor’. It also combines with stems of other origins, e.g. adikuasa (kuasa ‘power’) ‘superpower’ (the stem is also borrowed from Sanskrit, but this noun was formed in Indonesian).

antar‑ ‘inter’, according to Verhaar (1984:20–21) an affix of Sanskrit origin which has become productive according to Jones (1984:7). It may be a free form rather than a prefix, e.g. antar suku ‘intertribal’, antar kelompok ‘inter‑group’.

catur‑ ‘four’, according to Verhaar (1984:20–21) an affix of Sanskrit origin, e.g. caturwulan (wulan ‘month’ is Javanese) ‘four monthly’.

dwi‑ ‘two’, according to Verhaar (1984:20–21) an affix of Sanskrit origin, e.g. dwihari (hari ‘day’) ‘lasting two days, two-day’, dwirumah (rumah ‘house’) ‘dioecy’. It also combines with stems of non-native non-Sanskrit origin, showing that the prefix is productive, e.g. dwifungsi (fungsi ‘function’, from Dutch) ‘bifunctional’.

eka‑ ‘one’, according to Verhaar (1984:20–21) an affix of Sanskrit origin, e.g. eka bahasa ‘monolingual’ – the stem is Sanskrit, but the derivation was coined in Indonesian.

maha‑ ‘very (important person)’, according to Verhaar (1984:20–21) an affix of Sanskrit origin, e.g. mahabesar (besar ‘big’) ‘almighty’. Aside from its use as a prefix, this form apparently also exists as a free form, e.g. maha pelit (pelit ‘stingy’) ‘very stingy’ (examples provided by Uri Tadmor, personal communication, 2011).

nir‑ ‘privative, without, free from’, e.g. nirkabel (kabel ‘cable’, stem probably Dutch-derived) ‘wireless’. Has become productive according to Jones (1984:7). Another example is nirlaba (laba ‘profit’, Sanskrit-derived stem) ‘non‑profit’, which was likely coined in Indonesian.

pasca‑ ‘post-, after’, e.g. pascapanen (panen ‘harvest’) ‘post-harvest’, pascamakan (makan ‘to eat’) ‘post-meal tea’. Also, there is at least one formation with a Sanskrit stem, though it was likely coined in Indonesian, e.g. pasca sarjana ‘postgraduate’.

pra‑ ‘pre’, only a few words are formed with this and other affixes, which are part of the learned vocabulary, according to Mueller (2007:1220). According to Verhaar (1984:20–21), it is an affix of Sanskrit origin and has become productive on native stems (Jones 1984:7), e.g. prabayar ‘prepaid’ (among others), and on Sanskrit-derived stems, e.g. pra sarjana ‘undergraduate’.

purna‑ ‘after’, according to Verhaar (1984:20–21) an affix of Sanskrit origin, e.g. purna jual (jual ‘sale’) ‘after sales service’, purnajabatan (jabatan ‘position, post, function’) ‘previous position’.

sapta‑ ‘seven’, according to Verhaar (1984:20–21) an affix of Sanskrit origin, e.g. sapta pesona (pesona ‘charm’ of uncertain origin) ‘having seven charms’.

serba‑ ‘various’, according to Verhaar (1984:20–21) an affix of Sanskrit origin, which has become productive according to Jones (1984:7), e.g. serba ada (ada ‘to have, exist’) ‘having everything’, serba salah (salah ‘wrong’) ‘always wrong’ (examples provided by Uri Tadmor, personal communication, 2011).

swa‑ ‘self’, according to Verhaar (1984:20–21) an affix of Sanskrit origin, which has become productive according to Jones (1984:7), e.g. swabakar (bakar ‘to burn’) ‘self‑immolate’, swalayan (layan(an) ‘service’) ‘supermarket, literally self‑service’.

tata‑ ‘order, arrangement’, has become productive according to Jones (1984:7), being attested with native stems e.g., tata ruang ‘interior’. It is also found with other foreign non-Sanskrit stems, e.g. tata tertib ‘law and order’, and likewise with Sanskrit-derived stems in what seem to be neologisms coined in Indonesian, e.g. tata boga ‘table manners’.

tri‑ ‘three’, e.g. triwulan (wulan ‘month’ from Javanese) ‘three monthly’, triwarna (warna ‘color’ from Sanskrit) ‘three colors’. According to Verhaar (1984:20–21) this affix is of Sanskrit origin.

tuna‑ ‘struck, hurt, cut’, according to Verhaar (1984:20–21) an affix of Sanskrit origin, e.g. tunarungu (rungu ‘to hear’, from Javanese) ‘deaf’ (example provided by Uri Tadmor, personal communication, 2011).

2 agent or profession noun derivational suffixes

‑wan ‘masculine agent or profession noun’, reasonably common according to Mueller (2007:1220) and Jones (1984:7), it is also mentioned by Tadmor (2009:705), e.g. peragawan (peraga ‘display’) ‘male model’, rimbawan (rimba ‘jungle’) ‘forester’. It is also attested with non-native non-Sanskrit stems, e.g. kamerawan (kamera ‘camera’) ‘male cameraman’. According to Mueller (2007:1220), ‑man is an allomorph of ‑wan.

‑wati ‘feminine agent or profession noun’, reasonably common according to Mueller (2007:1220) and Jones (1984:7), also mentioned by Tadmor (2009:705), e.g. peragawati ‘female model’, kamerawati ‘female cameraman’.

Information and examples are from Malchukov (2006), who cites Myreeva (1964), see also Malchukov (2003).

1 hypothetical mood marker

‑dax ‘hypothetical mood’ (this form is part of the complex marker ‑r.dax‑, a combination of the native Evenki present tense marker ‑r(a)‑ with the Sakha hypothetical mood in ‑tax‑), e.g. waa‑r.dag‑im ‘I must have killed’, waa‑r.dag‑a ‘he/she/it must have killed’

6 person subject markers (all 6 borrowed) used only in combination with the hypothetical mood marker

‑im ‘first singular’, e.g. waa‑r.dag‑im ‘I must have killed’, waa‑r.dag‑a ‘he/she/it must have killed’

‑iŋ ‘second singular’, e.g. waa‑r.dag‑iŋ ‘you (sg.) must have killed’

‑a ‘third singular’, e.g. waa‑r.dag‑a ‘he/she/it must have killed’

‑put ‘first plural’, e.g. waa‑r.dak‑put ‘we must have killed’

‑kit ‘second plural’, e.g. waa‑r.dak‑kit ‘you (pl) must have killed’, suu gorolli‑r.dak‑kit (you.nominative/far_away_go‑probability‑2pl) ‘you (pl) must have gone far away’

‑tara ‘third plural’, e.g. waa‑r.dak‑tara ‘they must have killed’

1 ordinal number‑forming suffix

‑s ‘ordinal numeral derivation’, which is used in addition to a native derivational marker ‑i(s), e.g. d’uur‑i‑s ‘second’ (from d’uur ‘two’)

Malchukov (personal communication 2011) notes that the hypothetical mood marker is a “recently grammaticalized auxiliary verb, which explains why it comes with its own inflectional morphology.”

Information and examples are from Dalton‑Puffer (1996), see also Ciszek (2008) and Palmer (2009). Thanks to Sarah Menzel for help with the analysis. There are 8 derivational suffixes that are used on at least some Germanic stems in Middle English already (there are 10 others that are only attested in combination with Romance stems in Middle English and therefore not counted here).

4 abstract noun formation

‑age ‘abstract noun’, e.g. bondage

‑erie ‘abstract noun’, e.g. aldermanrie, husbondrie, loselry, outlawerie

‑ite ‘abstract noun’, e.g. scantetee

‑ment ‘abstract noun’, e.g. garnement

3 concrete noun formation

-ard ‘concrete noun’, e.g. niggard, dotard

-esse ‘concrete noun’, e.g. hunteresse

‑our ‘concrete noun’, e.g. worshippour, harpour

1 adjectivizer

‑able ‘adjectivizer’, e.g. spekable, knowable

Information and examples are from Steinkrüger (2003), see also Steinkrüger & Seifart (2009) and Seifart (2015).

2 adjectivizer prefixes

ma‑ ‘adjectivizer’, e.g. ma‑pyédra ‘stony, full of stones’

maka‑ ‘adjectivizer, cause the state or event denoted by the root’, e.g. maka‑myédo ‘dreadful, scary’, makarisa ‘funny’

2 verbalizer prefixes

man‑ ‘verbalizer’, e.g. man‑amígo ‘to be(come) friends’

pa‑ ‘verbalizer’, e.g. pa‑alísto ‘test each other’s skills’

1 reciprocal/locative suffix, which in its reciprocal meaning is used in combination with a borrowed verbalizer prefix, and which is therefore considered to be syntagmatically related to that prefix

‑an, ‑han ‘reciprocal/locative’, e.g. man‑kwénto‑han ‘to tell each other’, tubu‑han ‘sugar cane field’

3 non‑interrelated affixes

‑hin ~ ‑in ‘adjectivizer’, e.g. myedú‑hin ‘easily frightened’, asmá‑hin ‘asthmatic’

paka‑ ‘nominalizer’, e.g. paka‑alísto ‘talent’

ika‑ ‘ordinal number formation’, e.g. ika‑dos ‘second’, ika‑síngko ‘fifth’.

Information and examples are from Appleyard (2007). Bilin has borrowed an entire set of object markers from Tigre: “Alone among the Agaw languages, Bilin also has a set of object suffix pronouns added to the verb. Both the construction and the form themselves are borrowed directly from Tigre even to the extent that as in Tigre the 2nd person distinguishes masculine and feminine, a contrast that is not made in Bilin elsewhere, neither in the independent pronoun nor in the subject marking of the verb” (Appleyard 2007:491).

8 borrowed object markers

‑law ‘first singular object’

‑ka ‘second singular masculine object’

‑ki ‘second singular feminine object’

‑lu ‘third singular masculine object’

‑la ‘third singular feminine object’

‑na ‘first plural object’

‑kum ‘second plural object’

‑lom ‘third plural object’

These forms are also attested in Reinisch (1882), who did not recognize them as borrowed though. Reinisch (1882:38) also gives forms that distinguish gender in the third person plural (masc. ‑lom, fem. ‑län), which do not appear in Hamde’s (1986) grammar as noted by Appleyard (2007). Hamde (1986:33, 49–54, 93–94) gives the same forms as Appleyard (2007), without, however, a clear explanation of their function and distribution. Hamde (1986:48–52) mentions that suffixes are borrowed from Tigre and notes that “we cannot escape the fact that Tigre is invading Bilin through such grammatical explainable aspects of the grammar” (Hamde 1986:51), without giving further details. Hamde (1986:51–52) mentions two “prefixes and phrases” that would be borrowed from Tigre, one would be a verbal prefix (although written separately) translated as “as, that, is as” (Hamde 1986:51), another is apparently a nominal derivational element, translated as “mismal, improper” (also written separately). Hamde (1986) dismisses both as incorrect forms. It seems that both can be treated as borrowed lexical elements or maybe as particles. Neither Appleyard (2007) nor Hamde (1986) give examples of hybrid formation. However, since these forms are apparently obligatory, inflectional object markers, and there is no indication that all transitive verbs would be borrowed from Tigre, it is safe to assume that they give rise to hybrid formations.

Information and examples are from Xhuvani and Çabej (1962), Demiraj (1988:122–123), Svane (1992:290), Orel (1998), Boretzky (2004:1647), and Johanna Nichols (personal communication 14.7.2011). Examples were compiled by Lisa Morgenroth, some additional translation by Suzana Dibrani.

5 gender or declension class differentiated suffixes that form masculine nouns and adjectives

‑ac [ats] ‘adjectivizer’ (Boretzky 2004:1647; Demiraj 1988:122), e.g. sqepac ‘beak‑shaped’ (from sqep ‘beak’), thumbac ‘metal tip of a goad’ (from thumb ‘thorn’), dorac ‘one‑handed’ (from dorë ‘hand’)

‑aç [atʃ] ‘nominal derivation’ (Boretzky 2004:1647; Svane 1992:290), e.g. biraç ‘wall niche next to the hearth’ (from birë ‘hole’), gjembaç ‘thistle, sharp spine’ (from gjemb ‘thorn’), kungullaç ‘unripe, immature’ (from kungull ‘marrow, pumpkin, bottle’)

‑ak [ak] ‘nominal derivation’ (Boretzky 2004:1647), e.g. dorak ‘handle’ (from dorë ‘hand’), fytak ‘small wooden jug with a spout’ (from fyt ‘throat, gullet’), pishak ‘thick pine tree, stump’ (from pishë ‘pine’)

‑ash ‘adjectivizer’ (Demiraj 1988:122), e.g. buzash ‘lips’ (from buzë ‘end, edge’), gjumash ‘sleepy’ (from gjumë ‘sleep’), jargash ‘slobbering, drooling’ (from jargë ‘spit, phlegm’), kromash ‘scabious’ (from kromë ‘scabies, dogrose’)

‑avec ‘agent nominalizer’ (Demiraj 1988:122), e.g. jargavec ‘slug, slobbering’ (from jargë ‘spit, phlegm’), ligavec ‘slug, sniveling coward’ (from lig ‘evil, cowardly’), qullavec ‘person who is slow and clumsy’ (from qull ‘porridge, get soaked, soggy’)

3 gender or declension class differentiated suffixes that form masculine (‑ik) and feminine (‑itsə) nouns and adjectives

‑icë [itsə] ‘nominal derivation’ (Boretzky 2004:1647; Svane 1992:290; Demiraj 1988:123), e.g. petkicë ‘small sized garment’ (from petk ‘clothes’), tulicë ‘tender meat’ (from tul ‘meat, flesh’), balticë ‘ground that gets muddy quickly’ (from baltë ‘swamp, dirt’)

‑içë ‘nominal derivation or adjectivizer’, e.g. hardhiçë ‘wall lizard’ (from hardhucë ‘lizard’), barkiç ‘potbellied’ (from bark ‘belly, abdomen’), guriçkë ‘small rock’ (from gur ‘stone, rock’)

‑ik [ik] ‘nominal derivation or adjectivizer’ (Boretzky 2004:1647), e.g. balik ‘animal with white spotted face or body’ (from balë ‘badger’), baltik ‘muddy, marshy’ (from baltë ‘swamp, dirt’), çapik ‘impudent, saucy’ (from çap ‘to chew’)

2 gender or declension class differentiated affixes that form masculine (‑iʃtə) and feminine (‑inə) nouns, often having to do with locations

‑inë [inə] ‘location noun derivation’ (Boretzky 2004:1647; Svane 1992:290; Demiraj 1988:123), e.g. baltinë ‘muddy ground’ (from baltë ‘swamp, dirt’), botinë ‘swampland’ (from botë ‘earth, world’), djerrinë ‘wilderness, fallow land’ (from djerr ‘to destroy’)

‑ishtë [iʃtə] ‘location noun derivation’ (Boretzky 2004:1647; Svane 1992:290; Demiraj 1988:123), e.g. ahishtë ‘beech forest’ (from ah ‘beech‑tree’), arrishtë ‘walnut grove’ (from arrë ‘walnut tree’), baltishtë ‘ground that gets muddy quickly’ (from baltë ‘swamp, dirt’)

3 other nominal suffixes

‑nik ‘nominal derivation’ (Svane 1992:290; Demiraj 1988:123), e.g. danik ‘special purpose room’ (from daj ‘to divide’), drithnik ‘granary’ (from drithë ‘grain’), kohënik ‘contemporary’ (from kohë ‘time’)

‑kë ‘nominal derivation’ (Svane 1992:290), e.g. vickë ‘perverseness’ (from vithe ‘crupper, ramp’), veshkë ‘kidney’ (from vesh ‘ear, vine’)

‑ar ‘nominal derivation’ (Xhuvani and Çabej 1962:18), e.g. grunar ‘granary’ (from grunë ‘wheat’), lumare ‘riverbanks’ (from lumë ‘river’)

1 verbalizer

‑it (Demiraj 1988:123), e.g. djersit ‘to sweat’ (from djersë ‘perspiration’), shoshit ‘to sift, screen, scan’ (from shosh ‘to sift’)

The description of morphological borrowing from Spanish into various Quechuan languages is from Muysken (2012), who argues explicitly that these 7 suffixes are “part of a cluster of related processes, processes which almost operate in paradigmatic opposition” (Muysken 2012:485). Examples are taken directly from the sources that Muysken (2012) cites for Cajamarca Quechua (Quesada Castillo 1976a; 1976b).

7 derivational suffixes

‑dor ‘agentive, occupation noun’, e.g. michidor ‘shepherd’, awador ‘tailor’, ampidor ‘medicine man’ (Quesada Castillo 1976b:102)

‑iru ‘agentive, occupation’, e.g. yamtiru ‘firewood gatherer’ (Quesada Castillo 1976a:150, see also Muysken 2012:485, 491)

‑itu ~ ‑ita ­~ -situ ~ -sita ‘diminutive’, e.g. pishqito ‘little bird’, warmisita ‘little woman’, wishita ‘little sheep’ (Quesada Castillo 1976b:105)

‑liju ‘adjectivizer/nominalizer’, e.g. mancha‑liju ‘easily scared’ (Quesada Castillo 1976a:60, see also Muysken 2012:487, 491)

‑linku ‘adjectivizer/nominalizer’, e.g. wañu‑inku ‘dying, moribund’ (Quesada Castillo 1976a:96, see also Muysken 2012:487, 491)

‑nyentu ~ ‑chintu ~ ‑lyentu ‘adjectivizer’, e.g. mallaqnyentu ‘hungry’, qeshyachintu ‘sickly’, iskilyentu ‘having lice’ (Quesada Castillo 1976b:103)

‑likido ‘characterizing derivation’, e.g. qallo‑likidu (tongue‑liquid) ‘liar, talker, gossiper’, qalla‑likidu (lazy‑liquid) ‘lazy’, usa‑likidu (louse‑liquid) ‘full of lice’ (Quesada Castillo 1976a:125, 168, 169, see also Muysken 2012:487, 491)

Information and examples are from Hannß (2008) and Katja Hannß (personal communication, 2011).

3 case markers (out of a paradigm of 9 case markers in Uru)

‑kama ‘terminative case’, e.g. uskatan‑kama (moment‑terminative) ‘in a moment, in an instant’ (Hannß 2008:173; citing Métraux 1935:101)

‑taki ‘benefactive case’, e.g. suni‑taki ‘for the people’ (Hannß 2008:176; citing Muysken 2005:73)

‑na ‘locative case’, e.g. kʷas‑na ‘into the water’ (Hannß 2008:167; citing Vellard 1967:4)

3 suffixes that are not interrelated

‑pača ‘intensifier’, e.g. tonx pača ‘even today’ (Hannß 2008:253; citing Métraux 1935:91, 101)

‑naka ‘plural’, e.g. suni‑naka ‘the Uru people’ (Hannß 2008:138; citing Vellard 1967:3), tusa‑naka ‘balsa rafts’ (Hannß 2008:168; citing Métraux 1935:99)

‑pini ‘affirmative clausal clitic’, e.g. kustak-pini-ki (brother-affirmative-topic) ‘the truly older brother’ (Hannß 2008:274; citing Vellard 1967:24), yuk‑pini ‘very many’ (Hannß 2008:274; citing Vellard 1967:16)

Information and examples are from Bolozky (1999). The following six affixes are identified by Uri Tadmor (personal communication, 2011) as productive, and as having entered Hebrew from Russian, although they may be ultimately French, and may also exist in English, such as ‑er (from French ‑aire). Bolozky (1999) studies in detail the productivity of these affixes (in dictionaries, corpora, and by means of a “productivity test” involving native speaker judgments), providing also information on the extent to which they are used on native stems. See also Cohen and Laks (2012), who describe some of the same borrowed affixes.

4 agent noun formation suffixes

‑nik ‘agent noun’, e.g. núdnik ‘pest’, kibúcnik ‘kibbutz member’, klumnik ‘good‑for‑nothing’ (Schwarzwald 1998:271)

‑tšik ‘diminutive; doer’ politírtšik ‘furniture polisher’ (from politum ‘polish’), xaltúrištšik ‘one who does side‑jobs’ (xaltúra ‘side‑job, non‑serious job’), katántšik ‘tiny’ (from katan ‘small’), šaméntšik ‘chubby’ (from samen ‘fat’)

‑ist ‘agent, activist’ (Schwarzwald 1998:271), see also (Bolozky 2007:299–300)

‑er ‘agent noun formation’, e.g. širyoner ‘soldier in armor’ (from širyon ‘armor’) (Bolozky 1999:199)

2 abstract noun formation suffixes

‑ológya (according to Uri Tadmor only comically, such as xupológya ‘marital‑canopy studies’, xoxmológya ‘smart‑alec‑ness’)

‑izm, e.g. bitxonízm ‘emphasis on security’, bicu’izm ‘no‑nonsense, hands‑on approach’ (Bolozky 1999:111, 224)

Information and examples are drawn from Jones (1984), Verhaar (1984), Tadmor (2009), Torchia and Djuhari (2011) and personal communications with Uri Tadmor, David Gil, John Bowden, and a number of Indonesians working at the Jakarta Field Station of the Max Planck Institute for Evolutionary Anthropology in 2011, supplemented with etymological information from Wiktionary (https://en.wiktionary.org/ [accessed: 18.06.2024]) and internet searches. Only affixes that are attested in hybrid formations are given below. These include nouns with native Austronesian stems and borrowed stems that were likely borrowed before the commencement of Dutch influence.

5 miscellaneous nominal derivation, also used on adjectives

anti‑ ‘anti’, e.g. anti‑kuman (kuman ‘germ’) ‘anti‑germ’, anti nyamuk (nyamuk ‘mosquito’) ‘anti‑mosquito’, anti-cewek (cewek ‘girl’) ‘anti‑girls’, anti-cowok (cowok ‘boy’) ‘anti‑boys’, anti aturan jam malam (aturan jam malam ‘curfew’) ‘anti-curfew’. Much like in English, this affix can also be used as an independent morpheme modifying, for instance, a headless relative clause, as in saya anti pemikiran Tessa ‘I’m against what Tessa thinks’, Selebriti Anti Aturan Berkendara? ‘Celebrity against traffic regulations? (caption of a picture)’.

eks- ‘former’, e.g. eks pacar (pacar ‘romantic partner’) ‘ex-girlfriend/boyfriend’, eks tapol (tapol ‘political prisoner’) ‘former political prisoner’. Possibly, eks is an independent word.

kontra‑ ‘contra’, e.g. kontra pendidikan gratis (pendidikan gratis ‘free education’) ‘against free education’, also kontraalami (alami ‘natural’, Arabic stem) ‘unnatural’, saya kontra dengan pemikiran Tessa ‘I’m against what Tessa thinks’. Kontra may be an independent adverb rather than a prefix.

pro- ‘pro’, e.g. pro pendidikan seksualitas (pendidikan seksualitas ‘sex education’) ‘pro-sex education’, pro perkembangan (perkembangan ‘growth’) ‘pro‑growth’, kamu pro atau anti pemikiran Tessa? ‘Are you for or against what Tessa thinks?’ As indicated by the spelling, this may also be a free-form word.

sub‑ ‘sub-, noun denoting a subordinate version of a noun’, e.g. subbagian (bagian ‘part’) ‘subpart’, subpohon (pohon ‘tree’) ‘subtree’.

3 adjectivizing prefixes

inter‑ ‘inter-, forms adjectives from nouns’, e.g. intersuku (suku ‘tribe’) ‘inter‑tribal’, interkelompok (kelompok ‘group’) ‘inter‑group’.

intra- ‘intra-, forms adjectives from nouns’, e.g. intrakelompok (kelompok ‘group’) ‘intra-group’.

non‑ ‘negative, forms adjectives from nouns and predicates’, e.g. nonlingkungan (lingkungan ‘environment’) ‘non-environmental’, nonmigas (migas ‘fossil fuel’) ‘non‑fossil fuel’, nonmenarik (menarik ‘interesting’) ‘non-interesting’.

3 abstract noun forming suffixes

‑isme ‘abstract noun’, e.g. caloisme (from calo ‘agent or intermediary who negotiates government things on someone’s behalf’) ‘a mentality and system characterized by using the power of one’s position for personal gain’, goblokisme (neologism, from goblok ‘stupid’) ‘stupidity’, jahatisme (jahat ‘bad, evil’) ‘crime’, also premanisme (from preman ‘gangster’, a Dutch stem) ‘gangsterism’, a term most probably coined in Indonesian.

‑sasi ‘abstract noun’, e.g. kondomi‑sasi ‘spreading the use of condoms’, this example is also given by Verhaar (1984:24). This suffix is probably not the same as ‑isasi. Its status as an affix is disputed by de Vries (1984), see also Mueller (2007:1220).

‑isasi ‘transformation’, e.g. swastanisasi (swasta ‘private business’) ‘privatization’, kapurisasi ‘adding lime (calcium)’; in both examples, the stems are also borrowed from Sanskrit, but these nouns were formed in Indonesian. The complete form of this suffix is ‑(n)isasi with n occurring after vowel‑final roots. Its status as a productive affix is argued for by de Vries (1984), see also Mueller (2007:1220). bodohisasi (neologism from bodoh ‘stupid’) ‘becoming stupid, stupidification’, bungkamisasi (bungkam ‘to hush or gag someone’, Torchia and Djuhari 2011:84) ‘efforts to keep someone mum about a scandal’.

Information and examples are from Dawkins (1916) and Janse (2009a; 2009b), see also Gardani (2008:58–60). Cappadocian Greek has undergone heavy lexical and structural influence from Turkish, moving away from fusional and towards agglutinating structures, with some morphological borrowing. Different forms are attested in different dialects with only partial overlap. Dawkins (1916:130) explicitly notes that ‑lan ‘verbalizer, deriving intransitive verbs from adjectives’ and ‑t ‘causative’ “occasionally find their way into Greek words”.

1 verbal derivational marker

‑lan ‘verbalizer, deriving intransitive verbs from adjectives’, recorded in the variety of Ferték, e.g. ἀστενάρλανσε ‘he became ill’

1 valency-changing suffix

‑t ‘causative’, recorded in the variety of Ulağaç, e.g. ψοφάτσαν ‘they killed him’

2 subject agreement markers, in the varieties of Semenderé and (similarly) Sillí (Dawkins 1916:144)

‑k ‘first plural’, e.g. κέτουμιστικ ‘we exist’

‑inis ‘second plural’, e.g. κέτουστινις ‘they exist’

The borrowing of these forms is “probably due to the resemblance of ‑misti with the Turkish pluperfect in ‑mIs‑tI to which the 1st and 2nd person plural markers ‑k and ‑nIz are added. The Semenderé forms seem to replicate the Central Anatolian Turkish pluperfect in ‑DI‑mIș‑tI, e.g. ol‑du‑muș‑tu‑k → cé‑tun‑mis‑ti‑c” (Janse 2009b).

1 element in possessive pronouns

‑ï, a re‑analyzed Turkish element used in the formation of possessive suffixes, which are accordingly all partially Turkish, e.g. 1sg ‑ïm (combination of Turkish possessive element ‑ï and the element m, which occurs in the native Greek and Turkish form), 2sg ‑ïs (combination of Turkish possessive element ‑ï and native Greek element s), 3sg ‑ït (combination of Turkish possessive element ‑ï and the native Greek element t), 1pl ‑ïmas (combination of Turkish possessive element ‑ï and native Greek mas), 2pl ‑ïsas (combination of Turkish possessive element ‑ï and native Greek sas), 3pl ‑ïtne (combination of Turkish possessive element ‑ï and ‑tne, which is of unclear origin) (examples from Janse 2009b).

Information and examples mainly from Borg (1985:125–126) and Kossmann (2008; 2011) (see also Newton 1964; Tsiapera 1964:125–126; Roth 1979; 2003). Cypriot Arabic is also known as Kormakiti (Arabic). It has very many Greek loanwords which retain their Greek morphology, but relatively little Greek morphology is used on native Arabic stems, with one major exception: a set of suffixes marking diminutive, which is further specified for gender and number. These are productively used on Arabic nouns.

1 diminutive marker

‑u ‘diminutive’

4 gender/number marking suffixes

‑i ‘masculine singular’

‑kkya ‘masculine plural’

‑a ‘feminine singular’

‑úes ‘feminine plural’

Examples of hybrids: payt‑u‑i ‘little house’, payt‑u‑kkya ‘little houses’ (from payt ‘house’ (masc.)), mišl‑u‑a ‘little ladle’, mišl‑u‑es ‘little ladles’ (from mišl‑e ‘ladle’ (fem.)).

Information and examples are from Souag (2010), see also Souag (2009; 2020). There are five borrowed affixes/fusional patterns.

2 comparative/superlative (conventionally termed the elative) markers

C1C2əC3 template imposed on triliteral adjectives ‘comparative’, e.g. gzəl ‘shorter’ (from a-gzāl ‘short’), zṭəf ‘blacker’ (from azəṭtạ̄f ‘black’), zwər ‘bigger’ (from a-zuwwar ‘big’) (Souag 2020:481)

‑hum ‘superlative’, e.g. zwərhum ‘the biggest’ (from a-z-uwwar ‘big’) (Souag 2010:159). Note that Souag (2010:38) speaks of “full productivity of the Arabic comparative/superlative template in Siwi.”

2 nominalizing root patterns

āCəCCēCī template ‘agent (habitual doer)’, e.g. aččēwī ‘glutton’ (from əčč ‘eat’), aruwwēlī ‘flee‑er’ (from rwəl ‘flee’), ajəllēwī ‘someone who swears a lot’ (from jəll ‘swear’) (Souag 2010:422)

lə‑/əl‑ + C1C2āC3ət template ‘deadjectival nominalizer’, e.g. ləmlālət ‘whiteness’ (from aməllāl ‘white’), zzṭāfət ‘blackness’ (from azəṭtạ̄f ‘black’), əzzwārət ‘bigness, size’ (from a-zuwwar ‘big’) (Souag 2010:161)

1 verbal derivational prefix

dā‑ ‘suggestive’, e.g. dā‑təġḍəb! ‘let her get angry!’, dā‑iẓnant! ‘let them [Algeria and Morocco] divide it!’ (Souag 2010:384). It probably comes from Classical Arabic imperative daʕ ‘leave, let’ (Souag 2010:379).

Additionally, there are two plural morphemes which are each only attested on one native Siwi (Berber) stem. Therefore, they are not counted here: it can be expected that Souag’s (2010) detailed study would have uncovered further hybrid formations if they existed: lə‑CCaCəC, e.g. lŭgrazən ‘dogs’ (from agŭrzni ‘dog’) (Souag 2010:90), lə‑kdew‑a ‘squashes’ (from lə‑kdew‑iyy‑at ‘squash’) (Souag 2010:88).

Information and examples are from Heath (1978a; 1980a; 1984). 5 affixes were borrowed.

3 noun class prefixes (out of a paradigm of 8 noun class prefixes in Warndarang) (Heath 1980a:22–25); nouns can occur without prefixes, but usually take one.

(ṛ)a‑ ‘indefinite/non‑human class I’ (also optionally used in place of a human prefix), e.g. ṛa‑waṛiyi ‘Aborigine(s)’, ṛa‑jamaļagu ‘tapsticks’, ra‑waluŋma ‘python’ (Heath 1980a:23, 25, 26)

wu‑ ‘non‑human class II’, e.g. wu‑bunjiyu ‘honey’, wu‑ndawaynu ‘words’, wu‑jambaguwnu ‘tobacco’ (Heath 1980a:25)

ma‑ ‘non‑human class III’, e.g. ma-ḍuŋal ‘spear’, ma‑gamba ‘pearl shell’, ma‑yimbiḍgu ‘cypress’ (Heath 1980a:25)

2 case markers (out of a paradigm of 5 case markers in Warndarang) (Heath 1980a:26–28)

‑wala ‘ablative’, e.g. wuyulbawala ‘from the tree’ (Heath 1980a:26), wunaya‑wala ‘from here’, awuni‑wala ‘from there’ (Heath 1980a:53)

‑miri ‘instrumental’, e.g. maḍuŋamiri ‘by means of spears’ (three examples attested in corpora, but only one example given in Heath 1980a:26–27).

Information and examples are mostly from Boretzky (1975a:265–270), some additional examples stem from Xhuvani and Çabej (1962). Information on the etymologies of stems comes from Orel (1998). According to Boretzky (1975a:265), there are many Turkish suffixes in Albanian, but only 34 are used to derive more than a few Albanian stems. Boretzky (1975a:265–270) nevertheless lists 5 borrowed suffixes and gives examples of combinations with native stems for each. Some of them may be restricted to spoken varieties.

4 nominalizers

-xhi/çi ‘profession derivation’, e.g. derraxhi ‘swineherd’ (from derre ‘pig’), djathëxhi ‘cheese maker’ (from djathë ‘cheese’), lëkuraxhi ‘skinner’ (from lëkurë ‘skin’). Boretzky (1975a:265–270) found about 40 derivations with this suffix, but expects that there are many more. Brian Joseph (personal communication, July 2011) confirms that this suffix is productively used. It also appears in Genesin and Matzinger (2016:3127). This agent noun suffix is frequently borrowed from Turkic across languages in Eurasia as discussed in the AfBo entries on Uzbek affixes in Northern Tajik, Turkish affixes in Iraqi Arabic, Turkic affixes in Middle Mongolic, Azeri affixes in Udi, and Central Asian Turkic affixes in Russian. More examples are given by Hadzhieva et al. (2012), including -džija, borrowed from Turkish into Bulgarian. This form is now productively used in Bulgarian (https://en.wiktionary.org/wiki/-джия#Bulgarian [accessed: 10.07.2024]).

-llëk ‘abstract noun nominalizer (from adverbs, and others)’, e.g. zjarrlëk ‘burning heat’ (from zjarr ‘fire’). According to Brian Joseph (personal communication July 2011) this suffix is not very productive.

-li/-lli ‘designation of residents’, e.g. vendali ‘inhabitant’ (from vend ‘place’), Elbasanlli ‘inhabitant of Elbasan’, Tiranalli ‘inhabitant of Tirana’.

-qar ‘nouns denoting a person having a quality that is denoted by the noun or verb from which it is derived’, e.g. nihmaçar ‘helper’ (from ndihmë ‘help’), mundqar ‘someone who earns his daily bread with effort’ (from mund ‘effort’).

1 adjectivizer

-çe ‘ethnic or regional terms, and other derivations’, e.g. Shqipëtarçe ‘Albanian (adj.)’ (from Shqipëtar ‘Albanian (person)’), vendçe ‘locally’ (from vend ‘place’), derrçe ‘pig-like’ (from derr ‘pig’). Brian Joseph (personal communication, July 2011) confirms that this suffix is productive.

1 plural marker

-lar ~ -llar ~ -lerë ‘masculine human plural’, e.g. mbretlerë ‘emperor’ (from mbret, originally from Latin), priʃtlerë ‘priests’ (from priʃt, originally from Latin), giyshllar ‘grandfathers’ (from giyish, from Proto Albanian). Examples are from Gardani (2008:72), who cites Fiedler (1977). This form is used only on a few words, probably most of them of Turkish origin, according to Brian Joseph (personal communication, July 2011).

Information and examples are from Taylor (1954; 1956; 1959; 1977), Taylor and Hoff (1980), Hoff (1986), de Pury (2001; 2005), and Escure (2004:45–46; 2012). See also Grant (2010).

1 possessive person‑marker prefix

i‑ ‘first singular possessor’, e.g. iuaku ‘my drink’, iúti ‘my share (of food etc.)’ (Taylor 1956:39)

1 collective (plural) suffix

‑gu ‘collective, plural’, e.g. níbirigu ‘my younger siblings’, numégegu ‘my (personal) belongings’, nibą́iagu ‘my grandchildren’, tibegu ‘her people’ (Taylor 1959:190–191)

Note that Escure (2004:45–46) discusses a number of further affixes of putative Carib origin, in particular a nominalizing suffix ‑un(i), and a large number of evidential particles (or suffixes) of which she discusses in particular ‑ti (hearsay), ‑na (uncertainty), and ‑me (deductive). However, no corresponding elements have been identified in Carib (Kalin’a, Galibi), the source language for Cariban material in Garifuna. For the evidential particles, Escure (2004:45–46) cites similar‑looking forms from Hixkaryana as source forms (Derbyshire 1999:53), but Hixkaryana is from a different branch of the Cariban family. Additionally, the Garifuna evidential particles are only attested in Escure’s (2004:45–46) material, and not mentioned by other sources. For the nominalizing suffix, Lokono/Arawak (the Arawakan language most closely related to Garifuna) ‑n (Pet 2011:22–24 and passim), seems a likely cognate, which means the form would be native.

Information and examples mainly from Bereczki (1979; 1993; 2002) and Kangasmaa‑Minn (1998), in addition to other sources, as specified below.

4 case markers, out of 11 Mari case markers. Note that the “boundaries between case suffixes, derivational suffixes, and enclitics are far from clear‑cut” (Kangasmaa-Minn 1998:220).

‑la ~ ‑lä ‘modal, lative, comparative’, e.g. mar‑la ojla ‘speaks Mari’, totar-la solə̂k ‘Tatar scarf’, jal muča‑ŝkə̂‑la kaja (village/end‑illocution‑lative/goes) ‘goes to the end of the village’ (Kangasmaa-Minn 1998:227), kiɗemla ‘as my hand’ (Bereczki 1979:65)

-sɘ̂(n) ‘causal case, used with participles’, e.g. kottəmašən šolten o ̇ mɘ̂l ‘I did not cook so they would not eat’ (Bereczki 1979:68). See also Kangasmaa‑Minn (1998:233, 244). Note that Bereczki (1979:68) considers this to be a causal case suffix, borrowed from a corresponding case suffix from Chuvash, the use of which is restricted to participles. Its status as a case suffix is confirmed by the fact that it can be replaced by a dative case marker.

‑len ‘distributive’, e.g. kečɘ̂len ‘every day’, almašlen ‘changing’ (Bereczki 1979:69)

‑j ‘vocative case (used with kinship terms)’, e.g. isaj ‘my older brother!, my uncle!’ (Bereczki 1993:511). This form may also be from Tatar.

2 valency‑changing verbal suffixes

‑n ‘reflexive’, e.g. wijnem ‘get up’ (from wiem ‘lift up’) (Bereczki 2002:99)

‑tar, ‑tär, ‑dar, ‑där ‘causative’, e.g. jomdara ‘he/she looses’ (from jomam ‘I disappear’) (Mägiste 1968:10; citing Beke 1911:134–136, 287–288, 301–302). Kangasmaa‑Minn (1998:235) gives the forms =tə̂ ~ =də̂, e.g. joškar=tə̂ ‘make red’, but does not explicitly state that it is borrowed. This form may also be from Tatar.

6 miscellaneous derivational affixes

‑mVš ~ ‑šV ‘ordinal numeral formation’, e.g. kumuso ‘third’, nilemise ‘fourth’ (Bereczki 2002:99; citing Budenz 1864:437–438), koγә̑ mśo ‘second’ (Bereczki 2002:99; citing Galkin 1964:108), kolә̑ šo ‘twentieth’ (Bereczki 2002:99). Note that Kangasmaa‑Minn (1998:233–234, 244) considers this to be “the same (or homophonous) suffix (with an m‑ element)” as =sɘ̂(n) ‘causal case, used with participles’, which is used in the formation of ordinal numerals, e.g. nɘ̂lɘ̂mše ‘forth’.

ala‑ ‘indefinite’, e.g. ala‑kö ‘someone’, ala‑mo ‘something’ ala‑kuze ‘somehow’ (Bereczki 2002:99)

=le ~ =lö ~ =lo ‘adjectivizer’, e.g. lüm=lö ‘famous’ (Kangasmaa-Minn 1998:244)

=rak ‘comparative, modal’, e.g. joškar=gə̂=rak ‘reddish’ (Kangasmaa-Minn 1998:234), ko ̇yo‑rak ‘bigger’ (Mägiste 1968:10; citing Beke 1911:134–136, 287–288, 301–302). See also Raun (1971:120 and passim) and Wiedemann (1847:50). This form may also be from Tatar.

=lə̂k ‘abstract noun’, e.g. kül‑eš=lə̂k ‘necessity’ (Kangasmaa-Minn 1998:244), porlək ‘goodness’ (Mägiste 1968:10; citing Beke 1911:134–136, 287–288, 301–302)

=kalə̂‑ ‘iterative aspect’, e.g. lüj=kalə̂‑ ‘shoots repeatedly’ (Kangasmaa-Minn 1998:245)

Note that it is not clear what the total number of cases is. Kangasmaa‑Minn (1998:226) gives 10 (in addition to unmarked nominative), to which we added -j ‘vocative’.

Information is drawn from Noonan (2003; 2008), for some background see also Noonan (1996). Examples are from Noonan and Bhulanja (2005).

4 case markers, out of 22 Chantyal case markers

‑siŋ, ‑səŋ, ‑səŋə, ‑səŋgə ‘allative/comitative’, e.g. na‑siŋ ‘I (had)’, gəjjab‑siŋ ‘with goodness’. All examples in the literature consulted have a comitative reading, but Noonan (2003:319) characterizes this form as “allative/comitative”.

‑ma, ‑wã ‘temporal’, e.g. bis bərsə‑ma (twenty/year‑temporal) ‘in twenty years’

‑bɦənda ‘comparative/temporal’, e.g. nakhu‑bɦənda ‘(bigger) than a dog’ (example from Noonan 2003:320). All examples in the literature consulted have a comparative reading, but Noonan (2003:319) characterizes this form as “comparative/temporal”.

‑dɦin, ‑dɦikin ‘comparative/temporal’, e.g. khawa‑dɦikin (coming‑temporal) ‘since I had come home’, pəyle‑dɦikin ‘since the first’. All examples in the literature consulted have a temporal reading, but Noonan (2003:319) characterizes this form as “comparative/temporal”.

2 subordinators, out of 4 Chantyal subordinators

-nu ‘infinitive’, e.g. thũ-nu la-gəy a-thũ (drink-infinitive/do-progressive/negative-drink) ‘she was about to drink, but didn’t’

-khiri, -khir, -khi ‘cotemporal’, e.g. ca-wa-khiri (eat-nominalization-cotemporal) ‘when we eat’, thaa ta-wa-khiri (knowledge/become-nominalization-simultaneous) ‘when they know’

In addition, Chantyal borrowed all numerals (cardinal and ordinal), with two classifier suffixes for human vs. nonhuman used on numerals ‘one’, ‘two’, ‘three’. These classifiers are not counted here because their use is restricted to borrowed stems.

Information and examples are from Heath (1978a; 1980b).

4 suffixes which are not interrelated with other borrowed affixes

‑kaʔ ‘kin‑term dyadic dual’, e.g. gaykay‑kaʔ ‘uncle and niece’, mu:mu‑kaʔ ‘mother’s father’s sister and brother’s daughter’s child’, wa:wa‑kaʔ ‘elder brother and younger sibling’ (Heath 1980b:25–26)

‑ʔmayʔ ‘negative’, suffixed to verbs or other constituents, e.g. wa:n‑i‑ʔmayʔ ṋi: (go‑future‑negative/you) ‘You will not go’, munaŋa‑ʔmayʔ ra (White‑negative/I) ‘I am not a White’ (Heath 1980b:101)

‑bukiʔ ‘only’, rarely used, and no examples available (Heath 1980b:93)

‑ʔwañjiʔ ‘semblative case’, e.g. bakara‑ʔwañjiʔ ‘like long‑necked turtles’ (Heath 1980b:42) (out of a total of 14 case suffixes)

Note that four more prefixes might be borrowed, but the directionality is unclear, therefore they are excluded here (see also the entry on Ngandi): two out of two derivational adverbializers (or “compounding elements”) (Heath 1980b:81): malk‑ ‘times’ and bala‑ ‘side’; and two out of three verbal derivation markers (called “comitative” by Heath) (Heath 1980b:79, 82–83): baṱa‑ ‘applicative’, forming (semantically) transitive verbs with comitative object from intransitive verbs, and ṛay‑/yay‑ ‘applicative marker’, forming transitive verbs with an object that denotes something transported from intransitive verbs of motion.

Information and examples are from Lawrence Reid (personal communication, January 2012).

3 non-interrelated forms

=to/=nto ‘future’, encliticized to predicate, used in addition to inherited, preverbal future auxiliary. Only occurs in speech of younger generation (as of early 21st century), e.g. mapan=ka=nto ‘you will go’

mang‑/nang‑ ‘completed aspect, intransitive’, replaced native ma{n,n,ng}‑ +deletion of initial consonant of stem, e.g. nang‑sulbar ‘(they) solved’

taga‑ ‘a person from’ derivation, replaced native i‑; could be from Ilokano or Tagalog, e.g. tagabaryo ‘someone from the neighborhood’

Information and examples are from Bulut (2005; 2007; 2014), which is partially based on material presented by Buluç (1975). The borrowed affixes are attested in various Turkic varieties belonging to the South Oghuz dialect group spoken in Iran and Iraq.

2 object pronoun enclitics. Only the borrowed second singular and the third plural enclitics are attested in corpora, although it is very likely that other object pronoun enclitics (with different person and number values) are also borrowed (Christiane Bulut, personal communication 2012).

‑it ‘second singular’, e.g. almæ âllæm‑it (apple/buy.aorist.1sg‑2sg.dative) ‘I will buy you an apple’ (Buluç 1975:183; Bulut 2007:174)

‑şan ‘third plural’, e.g. yæyipt‑şan (eat.perfect.3sg‑3pl.accusative) ‘he has eaten them’ (Buluç 1975:183; Bulut 2007:174)

3 non‑interrelated suffixes

‑tar ‘comparative’, e.g. çux‑dar (much/most‑comparative) ‘more’ (Bulut 2005:254), “the copied morpheme +tar displays no combinatorial restrictions; it combines with Turkic and Iranian adjectives alike” (Bulut 2005:253)

‑aka ~ ‑eke ~ ‑ækæ ‘definite, specific’ (from Southern Kurdish, Gorani), e.g. ušaġ‑ækæ‑le’ (child‑specific‑pl) ‘those children’, mincuġ‑ækæ‑re (bead‑specific‑ablative) ‘of those pearls’, šê’r‑eke‑sin‑ne (poem‑specific‑possessor‑ablative) ‘about that poem by him’. This suffix “attaches directly to Turkic noun stems and precedes plural, possessive or case suffixes” (Bulut 2005:254). See also Kossmann (2011), who cites a Sorani source form from Blau (1980:46, 63). This form is also borrowed into Southern Iraqi Turkman.

‑iş ‘also, even’, e.g. yėmēduviş ‘you have not even eaten’, ġēliş ‘come!’, ōḫiş ‘even an eye’ (Buluç 1975:182). This suffix “is in all probability a copy of the Kurdish enclitic [...] which in southern Kurdish may be suffixed to either a nominal or a verbal form” (Bulut 2007:175).

An additional borrowed suffix, ‑i ‘indefinite’ from Persian, “with similar function [as ‑aka]” (Bulut 2005:254), is “in all instances of our material [...] attached to copied noun phrases” (Bulut 2005:255), i.e. stems that are likewise borrowed from Persian. For this reason, this suffix is not included here, although there is some indirect evidence that “could prove that the copied unit +i is compatible with non‑Persian nouns” (Bulut 2005:255).

Information and examples are from Heath (1978a; 1978b).

2 case markers (out of 7 Ngandi case markers) (Heath 1978b:41–50)

‑ṱu ~ ‑ḓu ‘ergative/instrumental’, e.g. ‑mumbaʔ‑ḓu ‘with axes’, awaṭuʔṇayi‑ṱu ‘his dog [bit me]’

‑ku ‘genitive, dative, purposive’, e.g. ŋayi‑ku ‘mine’, mawaṇguraʔ‑gu ‘the bandicoot’s’

Note that four further affixes may have been borrowed, but the directionality of borrowing is not clear, therefore they are not counted here (see also the entry on Ritharngu): (i) 2 out of 3 derivational adverbializers (called “compounding” elements by Heath (1978b:121)): malk‑ ‘times’, bala‑ ‘side’; (ii) 2 out of 2 verbal derivation markers (called “comitative” by Heath (1978b:83)): ‑baṱa‑ ‘applicative’ (forming (semantically) transitive verbs with comitative object from intransitive verbs), ‑ṛi‑ ‘applicative’ (forming transitive verbs with an object that denotes something transported from intransitive verbs of motion).

Information and examples are drawn from various sources, as noted below. For additional discussion of some of the following examples, see Rozencvejg (1976:24), Dahmen (1989:455), Breu (1992; 1996:35–36), and Boretzky (2004:1648).

2 agreement suffixes, marking agreement with neuter nouns on adjectives. The singular suffix occurs in both varieties of Istro-Romanian (viz. Northern Istro-Romanin and Southern Istro-Romanian), whereas the agreement marker -a only occurs in Northern Istro-Romanian.

-o marking mass neuter singular agreement, e.g. Southern Istro-Romanian nov-o ‘new-mass.neuter.singular’) (from Latin novum) (Gardani 2020:275; Kovačec 1966:68; Loporcaro, Gardani and Giudici 2020:92).

-a marking collective neuter plural agreement, e.g. Northern Istro-Romanian (diːtsa) bur-a ‘(children) good-collective.neuter.plural’ (from Latin bonum); it occurs on class one adjectives, articles, personal pronouns and demonstratives (Loporcaro 2018:295; Loporcaro, Gardani and Giudici 2020:118).

7 derivational prefixes, forming to some extent aspectual pairs. Further prefixes are mentioned in some of the sources, but these are not attested in hybrid formations.

po- ‘perfective’, e.g. potorče ‘to spin (perfective)’ from torče ‘to spin (imperfective)’ (Hurren 1969:62; Kovačec 1971:125), posúže ‘to suck (perfective)’ from súže ‘to suck (imperfective)’ (Sala 1988:79).

za- ~ ze- ‘perfective’, e.g. zadurmi/zedurmi ‘to sleep (perfective)’ from durmi ‘to sleep (imperfective)’, zaplənče ‘to weep (perfective)’ from plənče ‘to weep (imperfective)’ (Hurren 1969:62; Kovačec 1971:125), zalatrɑ̊ bark (perfective/inchoative) ‘to begin to bark’ from latrɑ̊ ‘bark.imperfective’ (Gardani 2020:276; Sala 1988:79).

raz- ‘perfective’, e.g. razlegua ‘to bind (perfective)’ from legua ‘to bind (imperfective)’ (Hurren 1969:62; Kovačec 1971:125; Sala 1988:79).

s- ‘perfective’, e.g. skopei ‘to dig (perfective)’ from kopei ‘to dig (imperfective)’ (Hurren 1969:62; Kovačec 1971:125).

od- ‘perfective’, e.g. odlomi ‘to break (perfective)’ from lomi ‘to break (imperfective)’ (Hurren 1969:62; Kovačec 1971:125).

na- ~ ne- ‘perfective’, e.g. namâŋcɑ̊ ‘to eat (perfective)’ from mâncɑ̊ ‘to eat’ (Sala 1988:79).

pri- ‘perfective’, e.g. prifače ‘to finish changing, to do differently’ from fače ‘to do’ (Sala 1988:79).

Similar sets of Slavic aktionsart prefixes have been borrowed into other Daco-Romance languages: 9 aktionsart prefixes found in Megleno-Romanian are borrowed from Bulgarian. The prefix do- ‘attainment of the final point of motion or activity’ is found in Romanian varieties spoken in Serbian (Vlach Romanian), e.g. do-facu preverb-do:past.3.singular) ‘s/he finished doing something’. Note that in Vlach Romanian, unlike Serbian (the source language) and Istro-Romanian (Kovačec 1971:125), the prefix does not have a perfectivizing role. Its meaning is derivational, and the category of Slavic aspect was not introduced with the borrowing (Gardani, Arkadiev, and Amiridze 2015:7). The borrowing of aspectual/aktionsart preverbs is common in languages that have been in contact with Slavic (see Russian affixes in Lithuanian Romani, and Russian and Belarusian affixes in Lithuanian).

Gardani (2008:69; citing Pușcariu 1943:280) also mentions that the accusative marker -u would have replaced the inherited -ɑ̌ in just only fixed phrase involving a native Romance lexical base, viz. za ţiru ‘at dinner‘ (cf. Latin cēna).

Information is from Rankin (1988:643–644, citing Geoffrey Kimball, personal communication). No examples of hybrids are given. The identification of the affix as borrowed is based on similarities in form, the presence of cognate forms in other Siouan languages, and the absence of similar forms in other Muskogean languages.

1 numeral-forming suffix

‑nąną ‘distributive of numerals’

These borrowed affixes are reported for the variety spoken in Hasankeyf (Turkey). Information and examples are from Andreas Fink (personal communication), see also Diem (1971) and Jastrow (1978:31) for the n‑pattern.

2 non-interrelated patterns

n-pattern ‘plural’, e.g. hənne ‘they’ (instead of native Arabic həmmu)

a‑ə pattern ‘perfect verb stem formation’ (no examples of hybrids)

Information and examples are from Nordhoff (2013), Smith (2013) and personal communications from Sebastian Nordhoff and Ian Smith, 2012.

1 polite imperative marker

‑nga ‘imperative’, e.g. vii ‘come! (fam.)’ vs. viinga ‘come! (polite)’, kummenga ‘eat! (polite)’

1 conjunctive participle marker

=tu ‘participle’, e.g. aka noos aka uusha kampani‑pa daa=tu, aka jaa‑faya dreetu (that/1pl/that/Usha/company‑dative/give=perfective.participle/that/past‑make/right) ‘we gave that to the Usha company and repaired it’

Information and examples are from Evans (1995:38, 129).

1 derivational suffix

‑kuru ~ ‑wuru ‘proprietive’, e.g. wara‑wuran‑kuru‑ (mouth‑food‑proprietive) ‘having food in its mouth’, kurndu‑kunawuna‑wuru‑ (chest‑child‑proprietive) ‘having a child on her chest’ (examples from Round 2010:199)

Information and examples are from Jones (1984), Verhaar (1984), Tadmor (2009) and personal communications from Uri Tadmor. Jones (1984:12–17) discusses loanwords and some “Arabic phrases and bound morphemes; Arabic grammar”, but it is unclear whether these are used with native Indonesian stems. According to Uri Tadmor (personal communication), only one affix from Arabic is attested with non‑Arabic stems.

1 suffix

‑awi ‘adjectivizer’ (Tadmor 2009:705), e.g. gerejawi ‘related to church’

Another Arabic adjectivizer, ‑iah, occurs only on Arabic roots and on roots perceived as Arabic according to Uri Tadmor (personal communication).

Information and examples are taken from Borg (1994:57), Borg and Azzopardi‑Alexander (1997:280, 291), Saade (2020:454, 455), and other sources cited below. See also Gardani (2008:75; 2012:81).

3 singulative suffixes

‑ata ‘singulative’, e.g. xemxata ‘sunstroke’ (from xemx ‘sun’), ksuħata ‘(act of) snobbery’ (from ksuħa ‘snobbery’), as described in Borg and Azzopardi‑Alexander (1997:280).

‑u ‘singulative masculine’, e.g. wizzu ‘goose’, du:du ‘worm’, fu:lu ‘bean’, see Borg (1994:57).

‑a ‘singulative feminine’, e.g. du:da ‘worm’, fu:la ‘bean’ (Borg 1994:57). Borg and Azzopardi‑Alexander (1997:174, 245) also mention singulative ‑a, e.g. nemla ‘ant’ (from nemel ‘ants’), xirja ‘a purchase’ (from xiri ‘shopping’), but do not say whether this is borrowed.

3 nominalizers

‑ż (masculine), ‑ża (feminine) ‘inhabitant nouns and adjectives from place names and adjectives’, e.g. Karkariż, Karkariża ‘person from Birkirkara’, Ċiniż ‘Chinese’, Ingliż ‘English’.

-azz ‘nominal from adjective or noun’, e.g. sakranazz ‘drunkard’ (from sakran ‘drunk’), it can have a pejorative connotation following Saade (2020:454, 455), see also Drewes (1994:94).

-iżmu ‘deadjectival noun’, e.g. lagħqiżmu ‘servilism’ (from lagħaq ‘to be servile’), cf. Saade (2020:454, 455), see also Brincat and Mifsud (2015:3350).

1 adjectivizer

-uż ‘adjectivizer’, e.g. nkejjuż ‘annoying’ (from nkejja ‘vexation’), as shown in Saade (2020:455), cf. also Drewes (1994:93).

1 augmentative suffix

-un ‘augmentative’, e.g. darun ‘large house’ (from dar ‘house’), cf. Saade (2020:454, 455), see also Brincat and Mifsud (2015:3350).

Information and examples are from Peterson (2011:324) and John Peterson (personal communication, March 2012).

2 so-called “sequential converbs”/”conjunctive participles” (out of a paradigm of 3 forms in Kharia)

=ke, e.g. kaʔ kom=ki ɖhoʔ=ke (bow/arrow=pl grab=sequential) ‘having taken their bows and arrows, ...’

=ker (no examples available)

Note that Abbi (1995a:183; 1997:139–142; 2001:47,53) mentions a number of other borrowed affixes in Kharia, but according to John Peterson (personal communication, March 2012), it is not certain whether any of these can be used with native stems.

Information and examples are from Riionheimo (2002; 2010). Note that Finnish and Estonian are closely related.

2 tense (and person) markers

‑p ‘third person present’, e.g. maksa‑a‑p ‘it costs’, istu‑u‑p ‘he sits’, tullo‑o‑p ‘he comes’

‑si ‘past’, e.g. ve‑i‑si‑mme ‘we took’, osta‑si‑it ‘they bought it’, jouta‑si‑mme ‘we were in time’

Information and examples are from Alexander Adelaar (2012:149–150; see also Adelaar 2005; 2009; 2010).

2 non‑interrelated prefixes

ki‑ ~ tsi‑ ‘nominalizer, diminutive’, e.g. ki‑tranu‑trànu, tsi‑tranu‑trànu ‘doll house’ (from trànu ‘house’)

h‑ ~ hu(‑) ‘future tense’, e.g. h‑angàlatră pàuli ‘Paul will steal’ (vs. m‑angàlatră Pàuli ‘Paul steals’), h‑umè‑nă àzi ni vùla ‘Money will be given to her’ (vs. umè‑nă àzi ni vùla ‘Money is given to her’), hu tùnga rahampìtsu ìzi ‘He’ll arrive tomorrow’ (vs. tùnga izàu ìzi ‘He arrives today’)

Information and examples are from Alexander Adelaar (2010). The identification of these affixes as borrowed is based on sound changes.

4 non‑interrelated affixes

vua‑ ‘passive’, e.g. vua‑helukă ‘guilty’ (compare mana‑melukă ‘to condemn O’), vuampanga ‘accused’ (compare mi‑ampanga ‘to accuse O’) (example from Adelaar and Vérin 1995:401), vua‑suràtr‑o ‘to be written’, vua‑valì‑ko ‘answered by me’, vua‑lazà‑ku ‘said by me’ (examples from Adelaar 2012:144)

tafa‑ ‘accomplished act (verbal aspect)’ (no example of hybrid formations available)

ra‑ ‘honorific personal prefix’ (used in nouns with human reference, including kinship terms), e.g. ra‑fuȼi ‘term of address for an old lady’ (fuȼi ‘white’) (example from Adelaar and Vérin 1995:403–404)

ta‑ ‘ethnic or geographical group’ derivation (no example of hybrid formations available)

Information and examples are from Rose (2012). Detailed reconstruction allows determining that the plural marker was borrowed from an intermediate Proto‑Carib language into Proto Wayampi‑Emerillon‑Zo’é. The marker is a suffix in two daughter languages (Wayampi and Emerillon) and a free form in Zo’é.

1 plural suffix

‑kom (Emerillon), ‑kõ (Wayampi) ‘plural’, e.g. wãĩwĩ‑kom (women‑pl) ‘women’ (Emerillon), e‑mɛmɨ‑kũ (1sg‑child‑pl) ‘my children’ (Wayampi)

Information and examples are from Parker Jones (2009:784).

1 derivational suffix

‑kona ‘sports event’ (from English marathon), e.g. hele‑kona ‘walk‑athon’ (from hele ‘go’)

A number of other English affixes in Hawaiian are not attested with native stems, i.e. heko‑kalame ‘hecto‑gram’, keni‑kalame ‘centi‑gram’, polai‑posapahate ‘poly‑phosphate’.

Information and examples are from work by Michael Rießler and Rogier Blokland, as noted below. These forms are used on native stems according to Michael Rießler and Rogier Blokland (personal communication), although no examples of hybrids are given below.

3 pragmatic particles

=že ‘topic’ (no examples of hybrid formations) (Rießler 2009:392; see also Rießler 2007:238)

=dak ‘topic’ (the exact function of the source form is described as “the preceding linguistic expression, say A, represents information, say x, on which a certain thought is based” (Post 2006:495)) (no examples of hybrid formations) (Rießler, personal communication, February 2012)

‑ka ‘hortative’, e.g. vancla‑ka ‘(let him) go!’ (Blokland and Rießler 2011:15)

1 negative indefinite prefix

nie‑ ‘negative indefinite’, forming expressions like ‘nobody, nowhere, nothing’ from expressions like ‘who, where, what’ (no examples of hybrid formations) (Rießler 2007:237–238; see also Blokland and Rießler 2011:14–15)

1 agent nominalizer

‑nihk ~ ‑nɨhk ~ ‑nehk ‘agent nouns’, e.g. pastlesnɨhk ‘a thorny shrub’ (from pastel ‘sharp’), v(ɨnn)ājtnihk ‘winegrower’ (Blokland and Rießler 2011:15; citing Mägiste 1968:16–17).

Information and examples are from Haig (2007:173; see also Matras 2007:45). The borrowing occurs in the western dialects of Kurmanji.

1 clause‑final clitic conditional marker

=se ‘conditional’, e.g. eer bapirê min ewna ne‑girt‑ine cem xa=sa ew‑na di‑mir‑in (if/grandfather/I/they/not_take/to/reflexive=conditional/they/die) ‘If my grandfather had not taken them in they would have died’

Information and examples are from Hildebrandt (2004:82–83; see also Hildebrandt 2007:292). Note that the same classifier is also borrowed from Nepali to Baram (see Nepali affixes in Baram). Both cases of affix borrowing are included here because Baram and Manange are only distantly related and the borrowing events are assumed to be independent.

1 numeral classifier

‑tʰa ‘general numeral classifier’, used optionally for every kind of noun (the only classifier in Manange, the Nepali source form is the non‑human classifier, opposed to a human classifier), e.g. 4ŋi‑ŋtha kòla (two‑classifier/child) ‘two children’, 4shi‑ŋtha 1pʌle (one‑classifier/hand) ‘one hand’ (numbers in the transcription indicate tone).

Information and examples are from Weinreich (1958:378–380; see also Reershemius 2007). Weinreich (1958:378) notes that the following eight additional potential cases of borrowed affixes have “rather limited distribution in the lexicon [and] are limited to a relatively small number of words each”, therefore they are not counted here: ‑še ‘feminizer’, ‑čik ‘endearing’, and five ‘pejoratives’ ‑ak, ‑áč, ‑éts, and ‑l’ák, ‑nák. The remaining thirteen borrowed affixes, given in the following, are said to be “highly productive” (Weinreich 1958:378). Weinreich (1958:378) notes that there are “many others”, without giving further details.

3 feminine noun derivation

‑ke ‘feminine’, e.g. lérerke ‘woman teacher’

‑ixe ‘feminine form of animal and (pejorative) personal names’, e.g. léjb‑ixe ‘lioness, Leyb’s wife’

‑ńítse ‘feminine form of personalizing or agentive nouns’, e.g. ejšes‑íš‑ńítse ‘adulteress’

2 (pejorative) adjectivizers

‑ske, e.g. oriman‑ske ‘fit for a pauper’

‑(ev)ate, e.g. tam‑evate ‘dull’

3 noun‑noun derivation

-arné ‘place name mildly contemptuous’, e.g. xasid‑arné ‘gathering place for khasidim’

‑ńik ‘personalizing or agentive derivation’, e.g. jišuf‑ńik ‘village Jew’

‑úk ‘pejorative, especially for names of occupations’, e.g. šuster‑úk ‘contemptible cobbler’

3 endearing vocatives/diminutives

‑(e)ńu, e.g. kínd‑eńu ‘dear child’

‑ink(e), used with nouns, e.g. múm‑inke ‘dear aunt’, with adjectives, e.g. dos rójt‑ink‑e kíšele ‘the little red (endearing) pillow’, with verbs, e.g. šlóf‑inke‑en ‘to sleep (nursery‑talk)’, with adverbs, e.g. pamél‑ink‑es ‘slowly (endearing)’, and with at least one interjection, e.g. gváld‑ink‑es ‘heavens!’

‑ičk, used (also) with adjectives, e.g. der ált‑ičk‑er man ‘the old (endearing) man’

1 privative prefix, “incipiently productive” (Weinreich 1958:378) and syntagmatically related to the personalizing or agentive derivational suffix ‑ńik in the sense that it “seems to be limited to constructions with ‑ńik in pejorative vocabulary” (Weinreich 1958:378).

bez‑ ‘without’, e.g. bez‑buš‑ńik ‘shameless person’

1 verbalizer of interjections

‑ke‑ ‘verbalizer’, e.g. bom‑ke‑n ‘to say “bom”’

Information and examples are from Capidan (1925) and other sources as noted below.

2 person markers on verbs. Examples of these are from Capidan (1925:94, 159), translations and evidence for Romance origins of the stems comes from Pușcariu (1905:168, passim), Weinreich (1953:32), and Gardani (2008:67). These suffixes are added to forms that are already inflected for the respective category by native suffixes -u and -i, hence, e.g. -um, -ǎm replaced native -u, and -iş replaced native -i. Gardani (2008:67), citing Sandfeld (1938:59), Capidan (1940:91), and Puşcariu (1943:274), states that “the morphemes have been added to the corresponding Romanian morphemes -u and -i, but are not productive since they apply only to certain verbs”, implying that at least some verbs select them, as the examples below suggest. Friedman (2012:327) points out that an alternative internal explanation of the phenomenon – in terms of a Megleno-Romanian conjugational restructuring – is also plausible.

-m ‘first person singular indicative present’, e.g. aflum ‘I find’, antrum ‘I enter’, amnum ‘I go’.

-ş ‘second person singular indicative present’, e.g. afliş ‘you (sg.) find’, antriş ‘you (sg.) enter’.

1 privative particle

ni- ‘privative particle’, e.g. ni-bun (privative-good) ‘evil’, ni-ḁn-grup-at (privative-aspect-bury-participle) ‘unburied’ (Capidan 1925:197). The second example shows that this prefix is not in the same morphological slot as the aspectual/aktionsart prefixes. The source form in Old Bulgarian is a clitic that attaches to stems separately from aktionsart prefixes (e.g., ne-pri-čestuvanje [privative-aspect-communion] not receiving communion’) and does not convey any aspectual meaning. In Megleno-Romanian, ni- is not attested in finite verbs, only in participles and a few nouns and adjectives.

9 derivational prefixes (out of 13 in Megleno-Romanian). The ones listed here include pan-Slavic prefixes, but not prefixes that are exclusive to one Slavic language other than Bulgarian (e.g. only Serbian). Information and examples are from Capidan (1925:195–202). No examples of hybrid formations are provided by Capidan (1925:195–202), but some are explicitly described as “very productive”. The functions of these prefixes are hard to outline precisely, also in the Slavic source languages, they usually encode aspectual or aktionsart meanings and sometimes form pairs.

du- ‘aktionsart’ (from Bulgarian do-), described as “very productive”

iz- ‘aktionsart’ (from Bulgarian iz-)

nӑ- ‘aktionsart’ (from Bulgarian na-)

pri- ‘aktionsart’ (from Bulgarian prĕ-)

pru- ‘aktionsart’ (from Bulgarian pro-), described as “highly productive”

pu- ‘aktionsart’ (from Bulgarian po-)

pud- ~ put- ‘aktionsart’ (from Bulgarian pod-)

răz- ‘aktionsart’ (from Bulgarian răs-)

ză- ‘aktionsart’ (from Bulgarian za-)

Similar sets of Slavic aktionsart prefixes have been borrowed into other Daco-Romance languages: 8 aktionsart prefixes are borrowed from Croatian to Istro-Romanian. The prefix do- ‘attainment of the final point of motion or activity’ is found in Romanian varieties spoken in Serbian (Vlach Romanian), e.g. do-facu preverb-do:past.3.singular) ‘s/he finished doing something’. Note that in Vlach Romanian, unlike Serbian (the source language) and Istro-Romanian (Kovačec 1971:125), the prefix does not have a perfectivizing role. Its meaning is derivational, and the category of Slavic aspect was not introduced with the borrowing (Gardani, Arkadiev, and Amiridze 2015:7). The borrowing of aspectual/aktionsart preverbs is common in languages that have been in contact with Slavic (see Russian affixes in Lithuanian Romani, and Russian and Belarusian affixes in Lithuanian).

23 derivational suffixes are borrowed from Bulgarian. Below, forms are listed that Capidan (1925:186–195) marks as pan-Slavic, while those marked only as Serbian are excluded. Note that some forms are probably complex and others may be allomorphs. There are a total of about 50 derivational suffixes in Megleno-Romanian, many of the non-Slavic ones have etymologies in Latin (i.e. count as native), but there are also several borrowings from Greek and some from Turkish. No examples of hybrid formations are provided by Capidan (1925:195–202).

11 diminutives

-atš ‘diminutive’, e.g. ghiumatš ‘small brass pitcher’.

-aş ‘diminutive’ (from Slavic in general), e.g. cupilaş ‘little baby’.

-cǎ ‘feminine diminutive’ (from Slavic in general), e.g. izvorcǎ ‘little spring’, described as “very productive”.

-ic ‘diminutive’ (from Slavic in general), e.g. aric ‘yard (lit. little area)’.

-icǎ ‘diminutive’ (from Slavic in general), e.g. balitiicǎ ‘little lake’.

-itşcǎ ‘diminutive’ (from Bulgarian), e.g. cǎsitşcǎ ‘little house’.

-iṭǎ ‘diminutive’/’feminine from masculine nouns’ (from Slavic in general), e.g. gǎuriṭǎ ‘little hole’.

-oşcǎ ‘diminutive’, e.g. baroşcǎ ‘little puddle’.

-utş ‘diminutive’, e.g. fratutş ‘little brother’.

-uş ‘diminutive’, e.g. cǎldǎrǎş (sic!) ‘little bucket’.

-tšoc (-tše-oc) ‘diminutive’, e.g. cǎptšoc ‘small head’.

9 nominalizers and nominal derivation

-an ‘quality nouns’ (from Slavic in general and Bulgarian), e.g. mijlucan ‘the middle one (middle brother)’.

-ealǎ ‘abstract nouns from adjectives’, e.g. nigrealǎ ‘blackness’.

-ean ‘inhabitant of’ (from Slavic in general), e.g. cǎtunean ‘someone who lives in a hamlet’.

-eṭ (meaning unclear) (from Slavic in general), e.g. vărdăreṭ ‘wind (from Vardar?)’.

-i̯ǎ ‘nouns from adjectives’ (from Slavic in general), e.g. vuṭi̯ǎ ‘wealth’.

-ineṭ ‘inhabitant noun’ (from Bulgarian), e.g. Cupineṭ ‘someone from Cupa’.

-işti ‘place where x is/happens’, e.g. bǎnişti ‘bathing place’.

-niṭǎ ‘place name derivation’, e.g. valturniṭǎ ‘place where eagles nest’.

-utinǎ ‘collective nouns’, e.g. erbutinǎ ‘grassland’.

2 adjectivizers

-lif ‘adjectives from nouns’ (from Bulgarian), e.g. bǎnlif ‘wounded’.

-nic ‘adjective from nouns’ (from Bulgarian -nik) three attestations, counting one with -arnic, e.g. cǎvai̯nic ‘poor’.

Information and examples are mainly from Sasse (1985; 1991). According to Sasse (1985:91–92), borrowed affixes have become productive and can today be used to form hybrid formations from Albanian stems. Additional information is from Altimari (2011).

4 diminutive markers, out of a total of 7 diminutive markers in Arvanitic Albanian (Sasse 1991:235–239), which do not seem to be phonologically conditioned allomorphs.

-(á)ko ‘diminutive’, e.g. priftáko ‘little priest’, limited to a few cases. Sasse (1985) explicitly states that this form is borrowed.

-átʃ ‘diminutive’, e.g. kuðátʃ ‘little jar’, limited productivity. Sasse (1985) explicitly states that this form is borrowed.

-ác ‘diminutive (pup)’, e.g. ʎepurác ‘rabbit’s pup’. Its use as a diminutive with nouns other than animals is restricted to Greek stems (Sasse 1991:240).

-ópuł ‘diminutive (pup)’, e.g. uʎkópuł ‘wolf’s pup’, has limited productivity (Sasse 1991:241).

2 noun derivation suffixes, out of a large set of other noun-to-noun derivation markers, e.g. feminine occupation, etc. (Sasse 1991:241–246).

-jár ‘agent noun’, e.g. ʃurrjár ‘someone urinating’ (from ʃúrrə ‘urine’). This suffix is largely productive and partially native and partially borrowed from Greek (ι)άρης (Sasse 1991:242).

-jót ‘inhabitant of’, e.g. curkatjót ‘inhabitant of Kiurka’ (from curkát ‘Kiurka’). This suffix is productive and Sasse (1985) explicitly mentions that it is borrowed.

4 verbalizers, out of a total of 9 in Arvanitic Albanian. All verbal derivational morphemes are unproductive (Sasse 1991:252–255).

-(j)ás ‘verbalizer’, e.g. morrá ‘delouse’ (from mórr ‘louse’). The form is given as (j)ás (Sasse 1991:254) or á (Sasse 1985), but according to Sasse (personal communication, 2012), as is the correct form. Sasse (1985) explicitly mentions that it is borrowed.

-ís ‘verbalizer’, e.g. dzarrís ‘to harrow’ (from dzárrə ‘harrow’). It is only attested in a few words, yet Sasse (1985) explicitly mentioned that it is borrowed.

-ós ‘verbalizer’, e.g. ɲatrós ‘let oneself seduce’ (from ɲátrə ‘another’), only attested in one word. Sasse (1985) explicitly states that this form is borrowed.

-is-/-as-/-eks- ‘verbalizer’ to adapt loanwords from Italian that have a Greek etymology in Italian, as in the following examples from Altimari (2011:11): ngol-is-ënj (< incollare) ‘to paste’, compare Ancient Greek κόλλα (kólla) ‘glue’; shkerdhj-as-ënj (< scheggiare) ‘to chip’, from Latin schidia and ultimately Ancient Greek σχίδια (skhídia) ‘chip’; stirand-aks-ënj ~ stirand-as-ënj (< stiracchiare) ‘to stretch (washed clothes)’, probably from Ancient Greek τείρω (teírō) ‘to wear out, rub’, via Latin.

11 adjectives used as derivational prefixes or as bound first elements of compounds, out of a total of 11 such forms in Arvanitic Albanian. They are used on nouns, i.e. as adjective-noun compounds (Sasse 1991:257–258). All are productive and, for most, various examples with Albanian stems are found.

aγrio- ‘wild’, e.g. aγriodérr ‘wild boar’.

ðoʎo- ‘poor’, e.g. ðoʎonún ‘poor godfather’, only three attestations, all from songs.

meγalo- ‘big’, e.g. meγalokopíʎe ‘old spinster’.

mes- ‘medium’, e.g. mesnátə ‘midnight’.

mono- ‘single’, e.g. monobíʎə ‘single daughter’.

mavro- ‘black, pitiful’, e.g. mavropʎák ‘poor old man’.

paʎo- ‘bad’ (very popular), e.g. paʎodjáʎə ‘naughty child’.

psefto- ‘fake’, e.g. pseftopúnǝ ‘superficial work’.

proto- ‘first’ (only one example given), e.g. protohérǝ ‘for the first time’.

stravo- ‘crooked’ (only attested once), e.g. stravogrúa ‘stubborn woman’.

vromo- ‘dirty, smelly’, e.g. vromopúnǝ ‘dirty work’.

Very rarely and only in songs, Albanian adjectives are also used in such constructions (Sasse 1991:258).

2 prepositions as derivational devices on nouns, out of a total of 3 in Arvanitic Albanian (the native Albanian form, not counted here, is nənə ‘under’).

para- ‘next to’, e.g. paratátǝ ‘stepfather’.

pro- ‘before’, e.g. protátǝ ‘forefather’.

5 verbal derivational devices, as part of a systems of 8 Arvanitic Albanian ‘preverbs’; kata is also used in this function, but it is not counted here again (Sasse 1991:262–267). These forms have entered the language as part of Greek complex loanwords, but they have become productive with Albanian stems.

apo- (different functions, only few cases), e.g. apobǝ́nem ‘I become’, apohíng(ǝ)ra ‘I ate up’.

kse- (different functions, like German ent, aus, auf, very frequent, also with Albanian verbs), e.g. kseʎóð(əɲ) ‘let someone rest’.

ksana- ‘do x again’, e.g. ksanaháp(əɲ) ‘open again’, very productive, also with Albanian stems.

para- ‘excessively’, e.g. parahá ‘eat too much’.

kata- ‘intensifier’ (relatively rare), e.g. kataʎóðem ‘I become very tired, i.e. I work too hard’. This form is also used as a derivational prefix on adjectives, participles, and adverbs (as part of a system of 2 such forms in Arvanitic Albanian, along with the native Albanian form pa ‘without’), e.g. kata- ‘very’, e.g. katairi ‘brand new’ (from iri ‘new’), katanáni ‘right now’ (from náni ‘now’).

4 verbal derivational devices as first elements of compounds (Sasse 1991:267).

kalo- ‘good’, e.g. kalopagúan ‘pay well’.

kako- ‘bad’, e.g. kakopagúan ‘pay badly’.

miso- ‘medium’, e.g. misoháp(əɲ) ‘open partially’.

proto- ‘for the first time’, e.g. protovéte ‘go for the first time’.

Information and examples are from Tenser (2005:13, 34–35).

1 superlative prefix

naj‑ ‘superlative’, e.g. nai‑bar‑edvr (superlative‑big‑comparative) ‘the biggest’

5 aktionsart prefixes

po‑ ‘aktionsart’, e.g. po‑mang‑ ‘to ask’

pod‑ ‘aktionsart’, e.g. pod‑gij‑ ‘to approach’

ros‑ ‘aktionsart’, e.g. ros‑pxen‑ ‘to tell’

vy‑ ‘aktionsart’, e.g. vy‑pi‑ ‘to drink’

za‑ ‘aktionsart’, e.g. za‑pres‑ ‘to pay’

­Information and examples are from Gutiérrez‑Morales (2005; 2008). Note that the form of this affix is probably influenced by Nahuatl, where it appears as ‑teero in the complex loanword kuenteero (from Spanish cohetero) ‘fireworks specialist’, and where it is reinforced by similar Nahuatl form (Gutiérrez‑Morales 2008; Gutiérrez‑Morales 2005).

1 agent noun derivation

‑teero ‘agent noun’, e.g. kööp‑teeroj (firewood‑agent) ‘wood cutter’, nöts‑teeroj‑am (armadillo‑agent‑already) ‘armadillo hunter’, aa‑teeroj (canoe‑agent) ‘rower, paddler’, tö’öp‑teeroj (fish‑agent) ‘fisherman’, nö’‑teeroj (water‑agent) ‘water carrier’, muut‑teeroj (well‑agent) ‘well driller’, yooxa‑teeroj (work‑agent) ‘hard worker’

Information and examples are from Emeneau (1962) and Andronov (1980).

1 verbal prefix

a‑ ‘present‑future and imperfect’, e.g. ī‑a kāva, underlying form ī a‑kāva (I/future‑go) ‘I shall go’ (Emeneau 1962:56–57; Andronov 1980:68)

2 adjectival derivational suffixes

‑tir ‘comparative’, e.g. juāntir ‘better’ (from juān ‘good’) (Andronov 1980:45). This marker is originally Persian, but entered Brahui through Baluchi (Elfenbein 1998:398).

‑ēn ‘adjective formation’, e.g. pudēn ‘cold’, ‘anēn ‘sweet’, xarēn ‘bitter’, kubēn ‘heavy’ (Andronov 1980:29)

Information and examples are from Kruspe (2004:64–69, 81–85, 206–208). This is a case of extreme compartmentalization of borrowed and native morphology since all native morphology is non‑concatenative, and all borrowed morphology is concatenative. In this sense, all borrowed affixes are related. (For a typology of morphological compartmentalization, see Gardani 2021.)

2 valency‑changing prefixes

br‑ ‘middle voice’ (passivization of verbs, nominalization ‘have’ of nouns), e.g. br‑bɒy ‘be dug up’

p‑ ‘causative’ (used alternatively to non‑concatenative causative morpheme, with some roots only), e.g. p‑jʔjiʔ ‘to make dirty’

4 aspect prefixes, including one circumfix

tr‑ ‘happenstance’, ‘happen to x’ (used with roots or causative‑derived, or reduplicated roots from verbs), e.g. tr‑ca ‘happen to eat’

par‑ ~ pr‑ ‘excessive agent/performer’ (with verbs), e.g. par‑ca ‘one who eats incessantly, a glutton’

m(N)‑ ‘imperfective’ (derives intransitive verbs from nouns and imperfective verb forms from verbs), e.g. m‑nar‑deh ‘be denying’

b‑...‑an ‘collective’ (collective activity verbs from verbs), e.g. b<pa’loh>an ‘many people hiding together’

1 valency‑changing suffix

‑iʔ ‘applicative’ (increases valency, marks iterative aspect), e.g. glɔk‑iʔ ‘to laugh unkindly at someone’

1 nominalizing suffix

‑an ‘nominalizer’ (relatively infrequent), e.g. jʔjiʔ‑an ‘dirtiness, filth’

Information from Marc Donohue (personal communication, 2012). Although no examples of hybrid formation are available, this prefix seems to be used with native stems.

1 passive marker

di‑ ‘passive’, used as a passive in subordinate clauses only in Tukang Besi, in addition to native (cognate) markers i‑ and (less commonly used) ni‑. What additionally may have helped borrowing is that Tukang Besi past tense marker is di‑ ~ i‑, i.e. the same allomorphy as the resulting passive marking pattern (Donohue 1999).

An additional, marginal case is kempe‑, a reanalysis of the first two syllables of kempetai ‘Japanese secret police during the occupation’ and Tukang Besi ‑taʔi ‘feces’. Now, kempe‑ can be used with native stems such as ‑buta ‘vagina’, ‑lau ‘penis’, etc. (Marc Donohue, personal communication, 2012).

Information and examples are from Haspelmath (1993:107) and from Don Stilo (personal communication 2013).

3 nominalizers (out of a total of 8 nominalizers in Lezgian)

‑qʰan ‘agent nouns from nouns’, e.g. xpe‑qʰan ‘sheep herder’

‑gan ‘nouns denoting containers from nouns’, e.g. t’urar‑gan ‘box for spoons’

‑ban ‘nouns denoting people associated with the thing denoted by the base word’, e.g. wak’ar‑ban ‘swineherd’

Information and examples are from de Reuse (1994). There are many loanwords and very many borrowed adverbial and conjunctional particles, but only 3 bound forms were borrowed, 1 suffix and 2 clitics. They seem to have a similar distribution (de Reuse 1994:288) and similar meanings, therefore they are considered to be paradigmatically related.

3 emphatic enclitics/suffixes

‑am ‘emphatic’, e.g. maatenllaam ‘this time’ (from maaten ‘time’, =llu ‘and’), sanganllaam ‘why now’ (from sangan ‘why’, =llu ‘and’), qayuqham ‘how is it? how did it go?’ (from qayuq ‘how’) (de Reuse 1994:364). This form is called an affix, but follows other clitics, such as =llu ‘and’ (de Reuse 1994:288). See also Miyaoka (2012:1549–1550), who analyzes this form as a clitic.

=qun ‘affirmative, emphasis’, e.g. natenqun ‘how (could you think)?’ (from naten ‘how’), esghaqaghnakenqun ‘Look!’ (from esghagh ‘see’), tuugkaatqun ‘Harpoon heads!’ (from tugkagh ‘harpoon head’) (de Reuse 1994:270–276, 424)

=Vy ‘emphatic vocative’, e.g. angyaay ‘Hey, boat!’ (from angyagh ‘boat’), qakmaay ‘You out there!’ (from qakma ‘out’), kangsughtungaay ‘I cannot fit through!’ (from kangsugh ‘cannot fit through’) (de Reuse 1994:285–287, 424)

Information from Nicholas Evans (personal communication, February 2012).

‑nganji ‘dear’ (applied to a relative) (no example of hybrid formation available).

Information and examples are from Mishra (1996). Only affixes for which there are indications that they are used on native stems are given in the following.

4 adjectivizers. About these four, Mishra (1996:98) explicitly states that they are borrowed and productive, Mishra (1996:48–51) gives many other affixes, about which this is not explicitly stated.

‑ū ‘adjectivizer’, e.g. pīṭū ‘fatal’ (from pīṭ ‘to kill’)

‑hā ‘adjectivizer’, e.g. bhūthā ‘haunted’ (from bhūt ‘ghost’)

‑yā ‘adjectivizer’, e.g. banyā ‘wild’ (from ban ‘forest’)

‑al ‘adjectivizer’, e.g. ḍubal ‘drowned’ (from ḍub‑ ‘to drown’)

1 gender marker, assuming that the following forms are phonologically conditioned allomorphs

‑ī, ‑in, ‑āin, ‑nī ‘feminine’, e.g. āl-i ‘girl’ (vs. ālas ‘boy’) (Abbi 1995b:4; 2001:47; Gardani 2012:83). These are “directly borrowed from Indo‑Aryan. The suffixes are mostly added to the borrowed nouns” (Mishra 1996:96, see also 21-22), but not exclusively (see also Abbi 1997:142).

2 forms deriving special numerals. The “suffixes khēp and bahrī are used to form enumerative/proportional numerals [...] the suffix ‑ō forming adjectival constructions from numerals” (Mishra 1996:96–97). Examples of ‑khēp and ‑bahrī are given in Mishra (1996:32), where they are also written as suffixes. Numerals one to four are indigenous Kurux (Mishra 1996:30). There are examples of ‑ō and ‑khēp in combination with numerals lower than five, but all examples of ‑bahrī are with the borrowed numerals five or higher, therefore ‑bahrī is not included here.

‑khēp ‘enumerative/proportional numerals’, e.g. ōnd khēp ‘once’

‑ō ‘adjectives from numerals’, e.g. tin‑ō ‘all the three’

1 passive marker

‑r/‑tār ‘passive’, e.g. cōx‑tār/cōx‑r ‘to be plucked’ (from cōx ‘to pluck’), la’ō‑tār/la’ō‑r ‘to be beaten up’ (from la’ō ‘to beat someone up’). Abbi (1997:140) explicitly claims this form is borrowed from Indo‑Aryan. Mishra (1996:108) discusses this form and says that the passive arose from contact with Indo‑Aryan, but does not say explicitly that the form ‑r/‑tār would be borrowed.

Abbi (1997:140, 142) mentions that the two Indo‑Aryan “conjunctive participles” ‑ar and ‑ki would be used in urban Kurux. Because they are written as separate words by Abbi (1997:140), and also because no corresponding forms were found in Mishra (1996), they are not considered here.

Information and examples are from Göksel and Kerslake (2005), Vietze, Zenker, and Warnke (1975), Kubiyak (2004), Memoglu‑Süleymanoglu (2006), Nişanyan (2009), Zengin (2009), and personal communication from Jaklin Kornfilt, as well as Kornfilt (1997:463–465). During the Ottoman Empire (13th‑19th century), there was a heavy influence from Arabic (especially in religious styles) and Persian (especially in literary styles), particularly in higher and educated registers (Göksel and Kerslake 2005:xxv–xxviii). Since 1911 there is a movement to clear Turkish of Arabic and Persian influence and “words containing Arabic and Persian suffixes are increasingly giving way to their synonyms of Turkish origin” (Göksel and Kerslake 2005:59). The following examples were compiled with the help of Sabine Günther. Only affixes that are attested with Turkish stems are given, there are also Persian affixes that are only used with Persian stems.

6 noun‑forming suffixes (including one forming nouns and adjectives)

‑baz ‘agent or occupation noun’, e.g. madrabaz ‘swindler’ (from mandıra ‘small cheesery’), dilbaz ‘somebody who’s good with words’ (from dil ‘tongue, language’), davlumbaz ‘shrouding, chimney hood, paddle box’ (from davul ‘drum’), düzenbaz ‘cheater’ (from düzen ‘regulation, scheme’), oyunbaz ‘playful person’ (from oyun ‘play, game, jest, trick’), kuşbaz ‘raiser, trainer, or seller of birds’ (from kuş ‘bird’)

‑dan (various noun derivations), e.g. yağdan(lık) ‘lubricator’ (from yağ ‘oil’), sonradan ‘afterwards’ (from sonra ‘after, later’), yaradan ‘the Creator’ (from yara ‘injury’)

-(h)ane ‘place noun derivation’, e.g. balıkhane ‘fish market’ (from balık ‘fish’), boyahane ‘dyeing factory’ (from boya ‘dye’), yemekhane ‘dining room’ (from yemek ‘food’)

‑kar ‘a person associated with a profession or occupation and other derivations’, e.g. koçkar ‘a ram that was raised for fighting’ (from koç ‘ram’), buzulkar ‘fern’ (from buzul ‘glacier’)

‑zede ‘victims of a catastrophic event’, e.g. depremzede ‘earthquake victim’ (from deprem ‘earthquake’), sınavzede ‘student who suffers from exams’ (from sınav ‘test, examination’)

‑dar ‘agent noun derivation’, e.g. bayraktar ‘standard‑bearer’ (from bayrak ‘flag’), sancaktar ‘standard‑bearer’ (from sancak ‘flag’)

1 adjective‑forming suffix (plus one forming nouns and adjectives, see above)

‑vari ‘adjectivizer’, e.g. yengeçvari ‘crab‑like’ (from yengeç ‘crab’), Türkvari ‘Turk‑like’ (from Türk ‘Turk’). There are also many new formations based on non‑Turkish stems, e.g. gangstervari ‘gangsterlike’, Amerikanvari ‘American‑style’.

Note that none of the very many derivational suffixes attaching to verbs were borrowed and also no suffixes that attach to nominals to form verbs (Göksel and Kerslake 2005:53–57). Note also that the set of Persian affixes in Azari (closely related to Turkish) is different.

Information and examples are from Dehghani (2000). The affixes given in the following are described as “borrowed prefixes and suffixes which are used frequently in Azari and which can attach to some native stems”, unlike others, which “can only attach to borrowed stems” (Dehghani 2000:96–97). Dehghani (2000:87–96) provides a clear description of the morphology, allowing to see how many forms of the grammatical subsystems are borrowed affixes.

2 privative-possessive adjectivizer prefixes. These two function as equivalents for the native suffixes ‑sIz and ‑lI (Dehghani 2000:97)

ba‑ ‘having the quality specified by the noun’, e.g. baädäb ‘polite’ (from ädäb ‘politeness’), basavad ‘literate’ (from savad ‘literacy’)

bi‑ ‘privative’, e.g. biädäb ‘impolite’ (from ädäb ‘politeness’), bisavad ‘illiterate’ (from savad ‘literacy’)

7 noun‑forming derivational suffixes (out of 10 such forms, i.e. there are 3 native ones, not counting “rarely used derivational suffixes”, of which another 2 derive nouns from nouns)

‑ban ‘a person who guards or maintains N as his usual occupation’, e.g. bagban ‘gardener’ (from bag ‘garden’), jängäl‑ban ‘forester’ (from jängäl ‘forest’)

‑baz ‘the person whose occupation involves N’, e.g. gumarbaz ‘gambler’ (from gumar ‘gambling’), gušbaz ‘bird keeper’ (from guš ‘bird’)

‑čA ‘diminutive’, e.g. bagča ‘small garden’ (from bag ‘garden’), käläkča ‘small boat’ (from käläk ‘boat’)

‑dan ‘standard container for N’, e.g. gänddan ‘sugar bowl’ (from gänd ‘sugar’), güldan ‘flower pot’ (from gül ‘flower’)

‑dar ‘the person who owns N’, e.g. eldar ‘the head of tribe’ (from el ‘tribe’), puldar ‘rich’ (from pul ‘money’)

‑Istan ‘a place designed to contain N’, e.g. gäbiristan ‘cemetery’ (from gäbir ‘grave’), gülüstan ‘rose garden’ (from gül ‘flower’)

­‑saz ‘the person who is the maker or repairer of N as his usual occupation’, e.g. čıragsaz ‘lamp maker’ (from čırag ‘lamp’), sahatsaz ‘watch maker’ (from sahat ‘watch’)

1 adjective‑forming derivational suffix (out of 3 such forms, the only two native denominal adjectivizing suffixes are the two mentioned above that are being replaced by the borrowed prefixes, not counting one rarely used derivational suffix, which derives adjectives from nouns)

‑I ‘pertaining to N, having the quality of N’, e.g. bazari ‘commercial’ (from bazar ‘market’), tarixi ‘historical’ (from tarix ‘history’)

The information and examples given below are taken from Koul (2005:156–158), which is the most recent one of the three descriptions available to me. Koul (2005:156–158) describes morphological borrowing in a section explicitly about borrowed morphology, and the affixes given below are explicitly described as combining with Kashmiri stems, even though this is in contradiction to an earlier description (Wali and Koul 1996:265) in at least one case, as noted below. Bhat (1987) and Wali & Koul (1996) list a lot more Persian affixes than Koul (2005:156–158), but they are less explicit about whether the forms they give are used with Kashmiri stems.

5 derivational prefixes. It is uncertain whether the results are nouns, adjectives or may be both.

be‑ ‘without’, e.g. bepatsh ‘untrustworthy,’ bezuv ‘without life’, bevəti:r ‘without manners’. Wali & Koul (1996:265) confirm that this prefix is used with native stems.

ba:‑ ‘with’, e.g. ba:darɨm ‘with religious piety’

har‑ ‘every’, e.g. harkã:h ‘every one’, harja:yi ‘every place’

kam‑ ‘little’, e.g. kamza:n ‘a little acquainted person’

bar‑ ‘on’, e.g. barz’av ‘on the tip of the tongue’

Note that there is also bad‑ ‘bad’, e.g. badza:th ‘bad character’, but Wali & Koul (1996:265) state that this prefix is only used with Persian stems, therefore it is excluded here.

2 suffixes deriving agent nouns

‑gor ‘agent noun’, e.g. g’avangor ‘one who sings’, gindangor ‘one who plays’, natsangor ‘one who dances’

‑ci: ‘agent noun’, e.g. tabalci: ‘one who plays on tabla’, ə:phi:mci ‘one who is addicted to opium’

2 suffixes deriving abstract nouns

‑gi: ‘abstract noun’, e.g. ga:nɨgi: ‘the act of procurer’, badma:šɨgi: ‘the act of a rogue’

‑i: ‘abstract nouns derived from nouns of agency’, e.g. ma:sṭar ‘teacher’ vs. ma:sṭəri: ‘teachership’, aphsar ‘officer’ vs. aphsari: ‘officership’, da:kṭar ‘doctor’ vs. ḍa:kṭəri: ‘medical practice’, ci:n ‘China’ vs. ci:ni: ‘Chinese’

4 suffixes deriving adjectives from nouns

‑ba:z ‘adjectivizer’, e.g. do:khɨba:z ‘deceitful’, ca:lda:z ‘crafty’

‑mand ‘adjectivizer’, e.g. phə:ydɨmand ‘useful’, akɨlmand ‘wise’, do:latmand ‘rich’

‑za:dɨ ‘adjectivizer’, e.g. darza:dɨ ‘an offspring of Dars’

‑da:r ‘adjectivizer “with x”‘, e.g. ləṭ’da:r ‘with tail’, chiṭida:r ‘with prints’. Koul (2005:157) gives this as an example of a combination of a Kashmiri prefix and Persian stem, but this is apparently the same as ‑da:r described by Bhat (1987:88), who describes it as “added to nouns, common and productive” and gives examples such as “land lord, leader, honest” which suggest that ‑da:r might be glossed as ‘with’.

Information and examples are from Sauer (1967:171–188; see also Sauer 1963; Mägiste 1968:6). According to Sauer (1967:171–172), borrowed affixes are primarily productive, also with native stems.

7 nominal derivational suffixes. Sauer (1967) gives a total of 11 borrowed nominal derivational suffixes, but 4 of these are excluded here because they are not attested with native stems

‑ja ‘adjectivizer’, e.g. wŭ’rja ‘fat (of bears)’

‑ki̬m ~ ‑kem ~ ‑kim ~ ‑kam ~ ‑kȧm ‘diminutive (note that all examples are with adjectives)’, e.g. łełkam ‘smallish’

‑kis ‘attenuative, adjectivizer’, e.g. warkis ‘light red, reddish’

‑an ‘instrument noun’, e.g. šestän ‘pole for pushing a boat along’

‑as ‘nominalizer’, e.g. šărtas keʌ ‘belt’

‑tek, ‑tak ‘adjectivizer’, e.g. šitak ‘peaceful’

‑tem ~ ‑tim ~ ‑tam ~ ‑tȧm ‘negative adjective formation’, e.g. ŏččam ‘unintelligent, stupid’

2 comparative/superlative particles

‑čək ‘comparative’, e.g. ȧjčək ‘smaller’

‑mȧtȧ ‘superlative’, e.g. ȧjmȧtȧ ‘smallest’

6 verbal derivational suffixes, Sauer (1963:195–196) lists a total of 7, but one is excluded here because it is only attested with Komi‑Zyrian stems. Sauer (1963) does not give examples of hybrid formations, but explicitly states that these suffixes are added to native stems.

‑aś‑ ‘reflexive verbs from nouns’

‑e̬d‑ ‘causative verbs from nouns’

‑al‑ ‘intransitive verbs from nouns’

‑śi̬‑ ~ ‑źi̬‑ ~ ‑t’śi̬‑ ‘reflexive, intransitive’ from nouns

‑i̬ś‑ ‘reflexive, frequentive’ [for this form, the only example seems to be verb‑verb derivation with valency change, but I am assuming that it can have a verbalizing function, too.]

‑ki̬‑ ~ ‑gi̬‑ ‘affective verbs’

Mägiste (1968:6), citing Sauer (1963) also mentions that there are a total of seven nominal derivational suffixes, seven verbal derivational suffixes, and two nominal particles. Sauer (1967:171–172, 182–188) discusses affixes from Russian, Samoyedic, Wogulian, and Tatar, but these do not seem to have become productive.

Information and examples are from Mägiste (1968:7).

1 nominalizer

‑ri ‘agent noun, occupation noun’, e.g. pellkuri ‘coward’, petturi ‘cheater’

2 feminine derivational affixes

‑inna ‘female person’, e.g. tohtorinna ‘female doctor’

‑ska ‘female person’, e.g. seppäskä ‘tailor’s wife’

Various other affixes from Swedish are used in the slang of Helsinki, but their use is not clear.

Information and examples are from Metzger (1998). This form also exists in a number of other South American languages, but the borrowing scenario for the Eastern Tucanoan languages Carapana, Tatuyo, Wáimaja, and Yurutí is quite clear: no other Eastern Tucanoan languages have it, but it is old in Arawakan. Examples below are from Carapana.

1 multifunctional prefix

ka‑ ‘deictic specifier or marker of specificity’, e.g. ka‑pakɨ ‘the/that father (the one referred to previously)’, ka‑bero ‘after the/that time (the time referred to previously)’

Information and examples are from Segura Munguía and Etxebarria Ayesta (1996) and Hualde and Urbina (2003). Eliasson’s (2012) detailed discussion of Romance influence on Basque contains a list of borrowed affixes that is largely consistent with the set of borrowed derivational affixes given here. Information on the borrowed participle marker is from Eliasson (2012), who refers to Trask (1995). See also Haase (1992:48–51), who focuses on French and Gascon influence on the Basque variety of lower Navarra, and Mujika (1982).

2 diminutive suffixes

‑ila, ‑ilo ‘diminutive’, e.g. neskatila ‘little girl’ (from neska ‘girl’), leihatila ‘little window’ (from leiho ‘window’), andrakila ‘doll’ (from andra ‘woman’), astokilo ‘little donkey’ (from asto ‘donkey’), gizonilo ‘little man’ (from gizon ‘man’) (Hualde 2003a:331; see also Haase 1992:49; Segura Munguía and Etxebarria Ayesta 1996:84, 89)

‑(i)no, ‑(i)na, ‑(i)ño, ‑(i)ña, ‑ño ‘diminutive’, e.g. emekiñio ‘very softly’ (from emeki ‘softly’), batño ‘a little one’ (from bat ‘one’) (Hualde 2003a:331; Segura Munguía and Etxebarria Ayesta 1996:89)

2 augmentative suffixes

‑nda ‘diminutive female’, e.g. urdanda ~urdanga ‘sow’ (from urde ‘hog’), oilanda ‘young hen’ (from oilo ‘hen’) (Trask 2003:117; Hualde 2003a:331)

‑ote ‘augmentative’, e.g. lodikote ‘kind of fat’ (from lodi ‘fat, thick’), handikote ‘kind of big’ (from handi ‘big’, cf. Sp grandote) (Hualde 2003a:331)