*endo

‘into’

Language
Proto-Anatolian
Grammar
adv
ID
3444

1. Individual Anatolian Languages

1.1 Hieroglyphic Luwian

/a-tá/, a-tá

‘in(to)’

1.1.1Transmission

This lexeme is attested over 75 times in all regions. Attestations date from the second half of the 13th century BCE to the early 7th century BCE.

[A.P.]

1.1.2Forms

indecl. /anta/ a-tá TELL AHMAR 6, §30 (Tell Ahmar)
      KARKAMIŠ A2+3, §18 (Karkamiš)
      KARKAMIŠ A11b+c, §13 (Karkamiš)
      ADIYAMAN 1, §4 (Commagene)
      ANCOZ 7, §11 (Commagene)
      MALPINAR, §4 (Commagene)
    a+tá BABYLON 1, §10 (Aleppo)
      TELL TAYINAT 2, l.4, 12 §i (Amuq)
      HAMA 4, §10 (Hama)
      IZGIN 1, §4 (Malatya)
    a-ta SULTANHAN, §18 (Tabal)
      KARATEPE 1 (Hu.), §20 (Cilicia)
      KARKAMIŠ A26f, §2 (Karkamiš)
      KARKAMIŠ A6, §7 (Karkamiš)
      MARAŞ 2, §2 (Maraş)
      KULULU 4, §4 (Tabal)
      TÜNP 1, §8 (Karkamiš)
      İSKENDERUN, §4 (Maraş)
    a-ta-i ASSUR letter e, § 21 (Assur)
    a-tax TOPADA, §20 (Tabal)
    a-ta? YALBURT frag. E, §2a (Konya)

The above table gives a representative selection of attestations, excluding broken attestations.

Yet another attestation may perhaps be found in 〈á-tá〉 at KARAHÖYÜK, §11 (cf. e.g. Hawkins 2000a:290, 293), but its existence has been called into question by Kloekhorst 2004a:31, who analyses the string of signs as a possible verb instead: 〈á-tá-i(a)-*515-ha〉 of unknown meaning. Note that other attestations are attributed to a different lexeme, i.e. /anta/ ‘also’.

[A.P.]

1.1.3Graphic Features

1.1.3.1Phonological Interpretation

The phonological interpretation of this word is certain, although the HLuw. spelling system prevents the pre-consonantal /n/ from being displayed in writing. However, the word is part of a larger family within HLuw. and beyond with CLuw. anta ‘in(to)’, Hittite anda ‘into’ etc. (thus noted already by Forrer 1932a:51; more recently see e.g. HED A:76f.). In fact, the frequent use of 〈tá〉 rather than 〈ta〉 with the present lexeme allowed Rieken 2010a:308f. to show that 〈tá〉 can be used to represent dentals in consonant clusters, particularly in the Karkamiš tradition.

The present lexeme is (probably) never spelt with 〈á〉 (sign *19) for its initial vowel (cf. Simon 2010b:252, 260), although it appears that some other members from the same word family are, see the adverb /antan/ ‘inside’, and the nouns /antaH(i)-/ ‘the inside(s), entrails (?)’ and /antatt(i)-/ ‘interior’.

With the word so frequently attested, it is possible to detect a move towards ligaturisation in some instances (thus already noted by Laroche 1960c:34), see e.g. MALPINAR, §4.

[AH.B.] [A.P.]

1.1.4Meaning

The general meaning of this adverb has long been clear, not least because of its obvious link with its Anatolian cognates, see Forrer 1932a:51 incl. n. 9 and 10Although Forrer translates it contextually as “zu” and “hinzu”. and later Laroche 1960c:34. In a statistical analysis, Boroday & Yakubovich 2018a:15 show that /anta/ can indicate a direction (‘into’) as well as a location (‘in’).

Like many other local adverbs, /anta/ can also be used as a postposition (with dative) and as a preverb, cf. e.g. Oshiro 1995a:214f. and García Trabazo 2019a:152, 155-158. The preverbal use, often lexicalised, can be observed e.g. in KARKAMIŠ A11b+c, §13: *a-wa/i-ta (SCALPRUM.PONERE2)u-pa-ní-zi a-tá (“PONERE2”)u-pa-ha, approx. /a=wa=tta uppaninzi anta ubahha/, translating as ‘I brought in the trophies’ (cf. Hawkins 2000a:103). In KARKAMIŠ A2+3, §7, /anta/ functions as a postposition: *a-wa/i-ta *a-mi-ia-za (LITUUS.AVIS)ta-ni-ia-za REGIO-ni-i a-tá (DEUS)BONUS-na (DEUS.VITIS)ti-PRAE-ia-ha ARHA (PONERE2)u-pa-ta, approx. /a=wa=tta ammiyanz taniyanz wattani anta Kumarman TiPariyan=ha ahha ubatta/, meaning ‘In my holdings, he (scil. the storm-god) brought forth (the gods) Kumarma and TiPariya in the area’ (cf. Hawkins 2000a:109).

[AH.B.]

1.1.5Stem

Indeclinable, this adverb is likely no longer morphologically transparent by HLuw. times, but has given rise to a number of derivations, e.g. /antan/ ‘inside’. Yakubovich 2010b:381 regards the word as a root cognate of /anni/ ‘under’.

[AH.B.]

1.2 Cuneiform Luwian

anda

‘in, into’

1.2.1Transmission

The word anda is well attested across the Luwian corpus.

[D.S.]

1.2.2Forms

indecl. an-da HT 1 ii 7’ (CTH 394, NS)
    KBo 8.17 obv. 8’ (CTH 208, NS)
    KBo 29.17 iii 3’ (CTH 767, NS)
    KBo 29.25 ii 3’ (CTH 764, MS)
    KBo 29.38, 13’ (CTH 769, MS?)
    KUB 9.31 ii 31 (CTH 757, NS)
    KUB 25.37 ii 20’ (CTH 771, NS)
    KUB 25.37 + KUB 35.132 ii 46’ (CTH 771, NS)
    KUB 35.107+ iii 21’ (CTH 764, MS)
    KUB 35.54 iii 19 (CTH 758, MS)
    KUB 35.8 i 7 (CTH 763, NS)
    KUB 35.51 ii? 27’ (CTH 760, NS)
    KBo 29.2 + KUB 35.52 ii 4´ (CTH 758, NS)
  a-an-da KUB 35.107+ iii 19’, 20’, 22’ (CTH 764, MS)
    KUB 35.54 ii 15’ (CTH 758, MS)
    KUB 35.65 iii 6’ (CTH 763, MS)
    KUB 35.89, 18’ (CTH 765, MS)
  an-˹da˺ KUB 35.104, 7’ (CTH 758, MS)
  [an-d]a KBo 29.25 ii 12’ (CTH 764, MS)
  an-ta KBo 13.263 r.col. 6’ (CTH 766, NS)
    KBo 29.32 iii 5’ (CTH 773, MS)
    KUB 35.45 ii 7’ (CTH 760, MS?)
    KUB 35.88 iii 12’ (CTH 766, NS)
    KUB 35.51 ii? 26’ (CTH 760, NS)
    KUB 35.51 ii? 28’ (CTH 760, NS)
  an-t[a] KUB 35.69, 2’ (CTH 761, MS)
  [a]n-ta KBo 19.155, 9’ (CTH 751, MS)
  ˹an-ta˺ KUB 35.129, 8’ (CTH 770, NS)
    KUB 35.54 ii 26’ (CTH 758, MS)
  a-an-ta KUB 35.101 obv.? 6’ (CTH 765, MS)
    KUB 35.108, 3’ (CTH 765, MS)
    KUB 35.108, 5’ (CTH 765, MS)
    KUB 35.108, 9’ (CTH 765, MS)
    KUB 35.108, 13’ (CTH 765, MS)
    KUB 35.69, 10’ (CTH 761, MS)
  an-˹ta-a(=)˺ KBo 25.54 + KBo 20.17 iv 13’ (CTH 665, OS)
  an-t[a] KBo 9.145 r.col. 2’ (CTH 770, NS)
  ˹an˺-t[a] KUB 35.54 ii 1’ (CTH 758, MS)

1.2.3Graphic Features

Checked against the photographs in the Konkordanz (Mainzer Fotoarchiv). It is interesting to notice that the scribe did not leave any space between ku-wa-at-ti and an-da at HT 1 ii 7’.

[D.S.]

1.2.4Meaning

The word anda was recognized early as the equivalent of Hittite anda ‘into’, although it often could not be translated accordingly since much of the remaining contexts were ill-understood (Otten 1953a:34, 56, Laroche 1958b:176, Laroche 1959a:28, Melchert in CLL:18, Yakubovich 2010d:269). Nonetheless, Laroche 1958b:176 adduces certain examples that fit perfectly the meaning obtained from the etymological perspective, e.g. anda wauwalipadda ‘(the wool) wrapped (it) in’ (KUB 25.39 iv 6), ÍD-i anda ‘(does not go) into the river’ (KUB 35.54 iii 19). As a pre- or postposition, anda governs the dative/locative (cf. the last cited example as well as HT 1 ii 7’ cited under Graphic Features).

The adverb anda is often attached to the sentence initial conjunction a, i.e. a=anta (e.g. KBo 9.145 r.col. 2’, KBo 13.263 r.col. 6’, KUB 25.37 + KUB 35.132 ii 46’, KUB 35.45 ii 7’), or it may even appear as the first tonic word of the clause, e.g. ānta=tta (KUB 35.51 ii 26’, 27’, 28’), for which Melchert (CLL:18) suggests the temporal meaning ‘in addition’ (cf. Hitt. anda=ma ‘in addition’), whereas Meriggi 1957b:68 proposed the disjunctive meaning ‘or’ for it. However, due to the poor understanding of certain contexts and their fragmentary states, it is difficult to ascertain any other grammatical function anda may have; but see under HLuw. for more information.

[D.S.]

1.2.5Derivatives

1.2.5.1Uncertain Derivatives

  1. anduwašša/i-

1.3 Lycian A

ñte

‘in, inside, also’

1.3.1Transmission

This word is well attested across the Lycian corpus.

[D.S.]

1.3.2Forms

indecl. ñte TL 57.9 (Antiphellos)
    TL 83.10 (Arneai)
    TL 72 (Kyaneai)
    N 337.3 (Limyra)
    N 337.11 (Limyra)
    TL 102.2(x2) (Limyra)
    TL 134.2(x2) (Limyra)
    TL 134.3 (Limyra)
    TL 87.4 (Myra)
    N 332.1 (Korba)
    TL 84.5 (Sura)
    N 333.11 (Tlos)
    TL 26.9 (Tlos)
    TL 75.2 (Tyberissos)
    N 320.16(x2) (Xanthos)
    N 324.16 (Xanthos)
    N 327.1 (Xanthos)
    TL 42a.2 (Xanthos)
    TL 44a.12 (Xanthos)
    TL 44b.39 (Xanthos)
    TL 44b.45 (Xanthos)
    TL 44c.19 (Xanthos)
    TL 45a.6 (Xanthos)
    TL 50.1 (Xanthos)
    N 306.2 (Ҫağman)
    N 306.3 (Ҫağman)
  ṇ̃ṭẹ TL 47.3 (Xanthos)
  ñt(e) TL 28.1 (Tlos)
    TL 51.1 (Xanthos)
  ñt[e] TL 26.1 (Tlos)
  [ñt]ẹ N 318.6 (Xanthos)
  ñta TL 31.3 (Kadyanda)
    TL 84.2 (Sura)
    TL 4.4 (Telmessos)
    N 324.22 (Xanthos)
    TL 36.3 (Xanthos)
    TL 38.7 (Xanthos)
    TL 39.7 (Xanthos)
    TL 41.4 (Xanthos)
  ñt[a] TL 5.6 (Telmessos)
  ñt TL 42b.2 (Xanthos)

The forms ñta are due to the vowel harmony triggered by the following verb, i.e. ñtatãtẽ (TL 31.3, TL 39.7), ñtatãti (TL 84.2), ñtatadẽ (TL 4.4, TL 36.3), ñtatuwa[xa] (N 324.22), ñtatetẽñtatãtẽ (TL 38.7), ñta=ija=tadẽ (TL 41.4).

The attestations showing the allomorphic form ñt- are due to syncope following univerbation, i.e. ñtadẽ (TL 42b.2) < *ñt-tadẽ < ñte tadẽ.

[D.S.]

1.3.3Graphic Features

For the damaged reading of ṇ̃ṭẹ (TL 47.3), see Christiansen 2020a:202. The word dde (TL 44b.29) has been suggested to represent a spelling variant of ñte, for which there is neither a contextual basis, nor is it independently supported by other Lycian data; cf. ddewe-.

[D.S.]

1.3.4Meaning

The word ñte was semantically elucidated very early on both contextual and comparative grounds with Greek and Latin by Deecke 1888a:136, an interpretation subsequently followed (Pedersen 1945a:23 comparing Hitt. anda ‘in’, Laroche 1958b:176, Meriggi 1980a:256, Melchert in DLL:45, Neumann 2007a:246, Christiansen 2020a:237 “with the verb ala(de)ha-(di)”). The word ñte can be used either as a postposition (e.g. ebei ñte ‘herein’ (TL 57.9), or simply as an adverb (e.g. N 320.16 “inside”), a preverb (especially with the verbs ta-(di) and tuwe-(ti)), or possibly also as an additive marker (e.g. TL 26.9, TL 44b.39). In any case, it is worth remembering that it is often impossible to distinguish between pre-/postposition and preverb with any degree of certainty.

[D.S.]

1.3.5Compounds

  1. ñtẽnẽ

  2. ñtada- (ñtele)

1.3.5.1Uncertain Compounds

  1. ñtipa-

1.3.6Derivatives

  1. ñtawãta-

1.3.6.1Uncertain Derivatives

  1. ñtere/i-

  2. ñtereze/i-

  3. ñterewete(/i)-

1.4 Lycian B

ñte

‘in, inside, also’

1.4.1Transmission

This lexeme is attested in the inscription of the inscribed pillar of Xanthos, dated contextually to shortly after the ascension of Artaxerxes II in 405-404 BCE and stylistically to around 400 BCE (Borchhardt & Eichner 1997-1999a:19f.).

[D.S.]

1.4.2Forms

indecl. ñte TL 44c.33 (Xanthos)
    TL 44d.22 (Xanthos)
  ñt(e) TL 44c.46 (Xanthos)

Pace DLL:123 and Schürr 2018b:74, ñt[.]nẽ at TL 44d.46-47 is here listed under the Lycian lexeme ñtẽnẽ. It is also found univerbated in Lyc. B ñtuwitẽni.

[D.S.]

1.4.3Meaning

The first instance at TL 44c.33 is likely to function as an additive marker such as “further(more)” (thus Shevoroshkin apud Neumann 2007a:246, Gehrisch 2018a:82 “also”, Sasseville 2021d:179 “further”; contra Schürr 2016d:166 “preposition”). The other example [.]utlade : ñte (TL 44d.22) is clearly a case of a postposition governing a dat./loc. Its function in TL 44c.46 is less clear, but it is perhaps a preverb (cf. Sasseville 2021d:179).

[D.S.]

1.4.4Compounds

  1. ñtada-

  2. ñtẽnẽ

1.4.5Derivatives

  1. ñtelija- (DN) (Sasseville 2021d:183)The word ñtelija is equated by Sasseville with the divine name Dantaliya- (cf. also URUantaliya), which occurs not only in Hittite but also in Luwian contexts. Comparing both the Luwian and Lycian evidence, it is possible to reconstruct the divine name as *endóli̯eh2- ‘the one belonging to the inside’. Needless to say, this etymological hypothesis cannot be proven until further information has surfaced regarding the personality of the deity. For a derivative of this same deity in Cuneiform Luwian, cf. Dantaliyašša- c., who helps with the water and the washing vessels in a birth ritual, see Sasseville 2018a:307.

1.5 Carian

δa (?)

‘in, under (in temporal sense)’

1.5.1Transmission

If it exists (see under Meaning), then hapax in a stele with a priest list (among others) from Hyllarima; the Carian heading is dated to 320 or 320/319 (Adiego 2019b:20–22).

[Zs.S.]

1.5.2Forms

indeclinable δa C.Hy 1,1

The attestation follows Adiego 2019b:18–20, 23 checked against the photograph in Laumonier 1934a:347 Fig. 26. Its existence is, however, doubtful since it entirely depends on the uncertain segmentation of an unintelligible sequence, see under Meaning.

[Zs.S.]

1.5.3Meaning ‘in, under (in temporal sense)’

The heading of the Hyllarima inscription consists of two lines, the second line of which together with the second half of the first line have been deciphered in Adiego 2019b (cf. also Simon 2021b):

1šasqarioδdymδa kduśo piĺipus usoτ 2muoτ armo trqδos qmoλš msoτ ylarmiτ
šasqarioδdymδa during the reign of Philippos, in the 4th year, in the month of Trqδ, the priests (of ?) the god (?) of Hyllarima’

Adiego 2019b:18–20, 23 cautiously proposed to segment δa from šasqarioδdymδa as a preposition preceding kduśo piĺipus ‘during the reign of Philippos (lit. during the Philippean reign)’, referring to the likelihood of the appearance of a preposition in this context as well as to potential cognates such as Carian δen ‘for’ (see s.v.) and Hittite anda, andan ‘in’. Nevertheless, he admitted that this is not necessary and left the case open (it was uncritically adopted in Yakubovich 2019c:303f.).

Considering that armo trqδos ‘in the month of Trqδ (lit. in the Trqδean month)’ has a completely identical structure and appears without a preposition δa, it is improbable that this δa should be segmented (unless ‘during the reign of X’ and ‘in the month of X’ was expressed differently, which cannot be excluded). Unfortunately, the initial sequence šasqarioδdymδa is currently completely opaque and thus, no falsifiable alternative can be provided. It should be, however, mentioned that a verb is still missing from the heading and °mδa exactly coincides with the 3pl. pret. form of the verb (i)m- (see s.v.) and, more generally, °δa exactly coincides with the 3pl. pret. ending (see Simon 2020d, also for more formal possibilities; it cannot be 3sg. contra Schürr 2022a, see Simon forthcoming). This issue must obviously be left open.

[Zs.S.]

1.6 Lydian

da-

‘(preverb)’

1.6.1Transmission

The preverb da- has been compared by Yakubovich 2005b to Hitt. and CLuw. anda. This hypothesis is made attractive by the comparison between da-cuwe- ‘to set up, dedicate’ and Lyc. n͂te tuwe- ‘to set up’ (Sasseville 2021a:384). The preverb da- is also found e.g. in dacal (see ca-(t)), dawicil (see wici-) and dakrod (see kro-(d)).

[D.S.]

1.7 Palaic

=anta

‘in, also, furthermore (?)’

1.7.1Transmission

This lexeme is found across the Palaic corpus.

[D.S.]

1.7.2Forms

indecl. an-ta KBo 48.178 iii 3’ (CTH 751, NS)
  an-ta KUB 35.165 obv. 3 (CTH 751, OS or MS)
  an-ta KUB 35.165 obv. 3 (CTH 751, OS or MS)
  an-ta KUB 35.165 obv. 8 (CTH 751, OS or MS)
  an-ta KUB 35.165 rev. 6’ (CTH 751, OS or MS)
  an-ta KUB 35.165 rev. 7’ (CTH 751, OS or MS)
  an-ta KUB 35.165 rev. 8’ (CTH 751, OS or MS)
  an-ta KUB 35.165 rev. 8’ (CTH 751, OS or MS)
  an-ta KUB 35.165 rev. 10’ (CTH 751, OS or MS)
  an-ta IBoT 2.35 + KBo 19.150 obv. 8 (CTH 751, MS)
  an-ta IBoT 2.36 obv. 4 (CTH 751, NS)
  an-ta IBoT 2.36 obv. 5 (CTH 751, NS)
  an-ta KUB 32.18 i 3’ (CTH 752, OS or MS)
  an-ta KUB 32.18 i 4’ (CTH 752, OS or MS)
  an-ta KUB 32.18 i 5’ (CTH 752, OS or MS)
  an-t(=) KUB 32.18 i 6’ (CTH 752, OS or MS)
  an-t(=) KUB 32.18 i 17’ (CTH 752, OS or MS)
  an-t(=) KUB 35.168, 5’ (CTH 752, OS)
  an-ta KBo 62.43, 3’ (CTH 751, NS)
  an-ta KBo 19.152 + KBo 27.77 iv! 15’ (CTH 751, MS)
  an-ta KUB 32.117+ ii 4 (CTH 752, OS)
  an-ta KUB 35.153, 3’ (CTH 754, NS)
  an-ta KBo 49.84, 3’ (CTH 754, NS)
  an-ta KBo 29.61, 7’ (CTH 754, NS)
  ⸢an-ta⸣[ CHDS 2.8, 6’ (CTH 754, NS)
  ⸢an-ta⸣[ KBo 49.84, 1’ (CTH 754, NS)
  a]n-⸢ta⸣[ KUB 35.153, 2’ (CTH 754, NS)
  [a]n-ta KBo 19.153 iii 10’ (CTH 751, NS)
  an-t[a] KBo 53.301 6’ (CTH 751, NS)
  a[n-ta] KBo 19.154 + KUB 35.166, 3’ (CTH 751, MS?)
  a[n-ta] KBo 19.154 + KUB 35.166, 12’ (CTH 751, MS?)
  an-da IBoT 2.35 + KBo 19.150 obv. 9 (CTH 751, MS)
  an-da KBo 49.84, 2’ (CTH 754, NS)
  an-da KBo 49.84, 6’ (CTH 754, NS)
  (=)n-ta KBo 19.155, 19’ (CTH 751, MS)

Note the elisions either with the initial or final vowel in [n]a-di-i-e=n-ta [ (KBo 19.155, 19’) and a=an-t=i-en-ta (KUB 35.168, 5’).

[D.S.]

1.7.3Meaning

From an etymological perspective, Kammenhuber 1959a:14 analyzed =anta as a preverb meaning ‘in(to)’. This seemed at first adequate in sentences, where the verb šuwa- ‘to fill’ occurs, i.e. ‘to fill in’, and is certainly correct in the sentence a=ant(a)=inta mārḫaš ‘The guests went in.’ (CTH 752). However, a consideration of all attestations shows that the meaning ‘in(to)’ does not work for most attestations.

One passage where a preverb with the meaning ‘in’ will not work is the following:

gāni=du=tta kuiš anitti maš=ta aniyaši aniya=ku=anta tabarni MUNUStawanani

‘Whoever accomplishes (things) in this way for you, as much as you will accomplish (for him), accomplish also for the king and queen!’ (KUB 35.165 rev. 10’)

One should note that the enclitic =anta occurs only in the last sentence. Another instance, where the meaning ‘in(to)’ does not work is the following one:

kuwāiš=a[=tt]a ḫalpūta ta[kk]uwānteš ašandu ānta kuwaniš tašūra takkuwanteš ašandu ānta warlaḫiš kuwalima takkuwanteš

“The bovines are displayed in the cowshed. The hounds are displayed in the kennels. The sheep (or lambs?) are displayed in the stockyard.” (KUB 35.165 rev. 6’-7’)

In the preceding passage, there are three parallel sentences. The enclitic =anta is used only in the second and third sentence. Therefore, the Palaic enclitic particle =anta seems to have acquired another function, which is interpreted as “additive” by Sasseville & Yakubovich 2018a:54f. “also, then” and paralleled by HLuw. /anta/ ‘also’ and Lycian B =ñte ‘in, inside, also’, which is provisionally adopted here. One should also compare the emphatic use of Hittite anda at the beginning of the sentence HW A:106f. For the syntactic analysis of Pal. =anta, cf. Meriggi 1963a:3-6, Carruba 1969b:24-26.

[D.S.]

1.8 Hittite

anda

‘into’

1.8.1Literature / Comments

HW 2nd ed. A:97-107

2. Reconstruction

2.1Reconstruction ‘into’

Hitt., Pal. and CLuw. anda, Lyc. n͂te have been compared to Lat. endo and reconstructed as PAnat. *endo (HEG A-K:33, HED A:76f., Melchert 1994a:105, EDHIL:185, Brosch 2014a:364-366). Kloekhorst (EDHIL:185, followed by Brosch ibid.) wishes to reconstruct a zero-grade anda based on Lyc. n͂te (and also OIr. and ‘there’; cf. LIPP II:159), although it was rightly pointed out by Melchert 1994a:135 that the in Lyc. n͂te is due to a late syncope and not to an inherited zero-grade. It is also worth adding that the preverb n͂te is possibly unaccented, as it often merges with the verb, a process that has fully happened in Lydian. Therefore, there is so far no evidence for a zero-grade in Anatolian. As suggested in PAnat. *endom, *endo was probably abstracted from *endom, after the grammaticalization process occurred (contra LIPP II:152, but cf. on page 153 n. 35, where Dunkel concedes for the Greek evidence that ἐνδο- ‘in’ (e.g. ἔνδοθι, ἐνδογενής) may have been back-formed from ἔνδον). This may have been motivated by other adverbs ending in *-o.

[D.S.]

Cross-references