mariḫš(i)-

The root of CLuw. mariḫš(i)- ‘fuzz’, maraḫšiwal(i)- ‘having fuzzes’ and the broken off maraḫaššal[ can be posited as /mraH-/ (with mute vowel in the first sign) going back to a Proto-Anatolian root *mreh₂- (**mreh₃- is excluded by the retention of the -ḫ- beside -š-!) with an s-extension. To explain the presence of both i- and a-vocalism, i.e. *mrēh₂s-/mreh₂s-, an acrostatic paradigm of an i-stem would be the most direct reconstruction (the weak stem being preserved in the derivative maraḫšiwal(i)-), although it would constitute, as far as we know, the first example of this type of inflection in an Anatolian i-stem (though not in Indo-European, see PIE lemma).An original lengthened grade (in either Hittite or Luwian) is required to explain the lack of coloring beside *h₂, i.e. with Eichner’s Law, as well as the apparent lenition of the laryngeal in maraḫaššal[, if it is correctly interpreted as /mrahassal…/ (instead of /mr̥Hassal…/), reflecting an original root noun *mrḗh₂-/mréh₂-. Therefore, the reconstructed pre-form *mrēh₂s-/ mreh₂s- + -i- ‘fuzz, loose thread’ must be considered provisional.There is another more complicated scenario that could be suggested on formal grounds, although each step lacks independent support: the s-enlargement can also be taken as an original s-stem with an acrostatic inflection ē/e, which would explain the i- beside the a-vocalism, i.e. *mrēh₂-/mreh₂- + -es- (for a reevaluation of acrostatic s-stems in Indo-European, see Höfler 2014a, according to whom the lengthened grade would have been secondarily introduced by analogy and a regular proterokinetic s-stem would have to be posited for the original formation). The original s-stem was then derived with a possessive o-suffix, i.e. *mrēh₂-s-ó- ‘having X’, subsequently substantivized as an i-stem, i.e. *mrēh₂-s-i- ‘fuzz’, though without the expected zero-grade of the root (on the derivational chain, see Höfler 2015b). This would, however, create a problem in the derivative maraḫšiwal(i)-, which does not show the lengthened grade of the root, but rather the regular e-grade or perhaps even a zero-grade in case it stands for /mr̥Hsiwal(i)-/. Moreover, it is doubtful that the i-stem derived from the possessive suffix would have retained both the lengthened and non-lengthened grade and, thus, we would be forced to assume that the original lengthened grade was shortened and eventually syncopated due to the extra syllables caused by the addition of further suffixes (a similar phenomenon is used to explain the Luwianism anšaššiwiš in Nikolaev 2021c). Moreover, considering the new etymology suggested below, a Schwebe Ablaut would be difficult to explain in an acrostatic s-stem. In sum, the first solution presented above is more economical.

[D.S.]

The root *merh₂- likely meant ‘grab, treat violently’ and ‘reduce to small pieces, crush’, which bears a resemblance to the reconstructed meaning of *melh₂- ‘to grind, crush’ (see CLuw. malḫu-). This similarity raises the possibility of semantic influence between the two roots In Indic, the two roots fell together, as mentioned above.. However, aligning such a meaning with Luwian mariḫš(i)- c. ‘fuzz, loose thread’ poses challenges. Nonetheless, it’s conceivable that the latter term might be rooted in a concept such as ‘small piece, smidgen, crumb, noil’ (further elaborated below).

To explain the term mariḫš(i)- c. ‘fuzz, loose thread, noil’, one could consider the following scenario: an s-stem *mérh₂-s- n. ‘crushing’ might have served as the foundation for a denominal adjective *mr̥h₂-s-ó- ‘crushing; crushed’. This adjective likely gave rise to Middle High German mursch, and German morsch, both meaning ‘rotten, brittle’ (presupposing a barytone accentuation). Further support for this adjective comes from its substantivizations. In the Germanic context, various place, river, and island names, such as Merseburg in Germany or the island Mors in Denmark, could potentially trace back to or contain the stems Proto-Germanic *múrsō-, *mérsō-, *merzá-. These stems, as proposed by Bichlmeier 2017a:49f., could have carried the sense of ‘(place/riverbed/shore) characterized by gravel, crushed stone’. Additionally, *mŕ̥h₂seh₂- with the meaning of ‘crushing instrument, crusher’ > *marasā, might have led to Latin marra f. ‘a sort of hoe for tearing up weeds’. Moreover, the passive/resultative connotation of ‘crushed stuff’ could have continued into Romance terms referring to ‘debris’ (see above).

An alternative method to substantivize thematic adjectives involved their transformation into i-stems. In certain archaic-looking forms, this derivational process seemingly generated acrostatic i-stems, recognizable by a new ablaut grade in the root syllable (see Nussbaum 1998a:523 for more details). For instance, consider *h₂ek̑-ro- ‘top, pointed’ (yielding Gk. ἄκρος ‘topmost’, OLith. ašras ‘sharp’) and its substantivization *h₂ók̑ri-/*h₂ék̑ri- (resulting in forms like Lat. ocris m. ‘craggy mountain’ Liv. Andr., Gk. ὄκρις m. or f. ‘jagged point’ Hp., Gk. ἄκρις, -ιος f. ‘hilltop’ Od., h. Cer., Ved. áśri- f. ‘sharp side, edge’ Br+; as per Nussbaum 1998a:523). Similarly, *u̯n̥k-ró- ‘curved’ (Ved. vakrá-) underwent transformation to *u̯ó/énkri- f. ‘curved thing’ (as in Ved. váṅkri- f. ‘rib’ RV+; according to Weiss 1996a:204f.). There is no inherent reason to presuppose that this i-substantivization process exclusively produced o/e-acrostatic, but no ē/e-acrostatic forms.

This interpretation holds potential for various formations. For instance, it could apply to Greek δῆρις ‘battle, ‘contest’ < *dḗri-, which could be seen as a substantivization of *de/oró- ‘split’ (cf. δορός ‘leather bag’; see Nussbaum 2014b:236). Another case is Young Avestan zāiri- (nom. sg. zāiriš, acc. sg. zāirīm, voc. sg. zāire, gen. sg. zarōiš) ‘yellow (one)’, and Vedic hári- (the generalized weak stem?) ‘yellow (one); yellow horse, steed’ < *g̑ʰḗlh₃i-. Here, the substantivization could stem from *g̑ʰl̥h₃-o- (as seen in Lith. žìlas ‘gray’, Latv. zìls ‘blue’) or *g̑ʰelh₃-o- (as in OIr. gel ‘bright, white, glowing’). Similarly, Tocharian B yel ‘worm’, and Toch. A pl. walyi < *u̯ēli- could be considered a substantivization of *u̯ol-ó- ‘turning’ (Old Norse valr ‘round’). For these acrostatic i-stems, see Widmer 2004b:50.

This provides a potential avenue for interpreting mariḫš(i)- c. ‘fuzz, loose thread, noil’ < *mrḗh₂si- as ‘small piece, smidgen, crumb, noil’, considering it as a substantivization of *mr̥h₂só- ‘crushed’ (cf. German morsch ‘rotten, brittle’, Proto-Romance *marra ‘debris’). The schwebeablaut in the substantivization is not without parallels. However, it’s important to emphasize that this explanation remains highly speculative.

[S.H.]

Hittite marriye/a-, marra- ‘(med.) to melt (down), to dissolve, to stew or cook until tender; to heat up(?), to bring to a boil(?)’, Luwian (Glossenkeilwort) mariḫš(i)- c. ‘fuzz, loose thread’.Cf., e.g., EDHIL:558.

Vedic mr̥ṇā́ti ‘crushes, smashes’, mr̥ṇáti ‘crushes, smashes; seizes’, mūrṇá- ‘crushed’, etc.Cf., e.g., EWAia I 319f. s.v. MAR^I1, MAR^I2; Werba 1997a:312f. The continuants of *merh₂- fell together with *melh₂- ‘crush, grind’.

Greek μάρναμαι ‘fight, do battle, quarrel’, μαραίνω, aor. ἐμάρᾱνα, mid. ἐμαρηνάμην, pass. ἐμαράνθην ‘quench, waste, wither’, pass. ‘die away, go slowly out, waste away’, (?) μάρμαρος, μάρμαρον ‘stone, piece of rock’, etc.Cf., e.g., GEW II:174 s.v. μαραίνω, II:177f. s.v. μάρναμαι, II:177 s.v. μάρμαρος, EDG:904 s.v. μαραίνω, 907 s.v. μάρναμαι, 907 s.v. μάρμαρος

Latin mortārium n. ‘mortar’, morētum n. ‘a country dish composed of garlic, rue, vinegar, oil, etc.’, marra f. ‘a sort of hoe for tearing up weeds’, Galician marra ‘hammer’, Proto-Romance *marra ‘debris’ (Latin mar, mara, marena ‘debris, rubble (of alpine brooks); stone heap’, Southern French marran ‘debris, rubbish’, etc.).Cf., e.g., WH II:112 s.v. morētum, EDLIL:390 s.v. mortārium; mortārium, if based on *mr̥to- ‘crushed’, seems to require an aniṭ root *mer-, though participles such as cultum from *k^u̯elh₁- ‘turn, etc.’ (: colere) indicate that these could be based secondarily on the laryngealless stem of the underlying verb, which presupposes a *merere, *morere, or *mernere ‘to crush’ in the not-so-long-ago prehistory of Latin; Höfler 2017a:21 (on marra), REW:439 Nr. 5369 (on *marra) and Nr. 5370 (on marra)

Old Irish marnaid, ·mairn ‘betray, deceive, confound’, subj. ·mera, pret. ·mert, mrath o, n., Middle Welsh brad, Old Breton brat ‘treachery, treason, betrayal’, OIr. meirle ‘theft, robbery’.Cf., e.g., Schumacher 2004a:477f., EDPC:257, Zair 2012a:75

Old Norse merja (pret. marða) ‘beat, crush; bruise’, Old High German maro, marawi and mur(u)wi ‘brittle; delicate; soft’, Old English mearu ‘id.’, Old Norse meyrr ‘id.’, morna ‘wither away’, etc.Cf., e.g., IEW:736, de Vries 1977a:385 s.v. merja, 386 s.v. meyrr, 393 s.v. morna

Serbo-Croatian mȑva ‘crumb; bit’; Old Church Slavonic iz-mrъmьrati ‘to clear (woodland)’, Old Russian -moromradi ‘gnaw, crumble’, Slovene brámor ‘mole cricket’ (< *mormorъ ‘gnawer’).Cf., e.g., IEW:736, Snoj 2003a s.v. brámor