Tocharian Alternation of m \ w \ p (Draft)

2025

If Uralic *wVN > *mVN within a syllable was optional, it would explain v- vs. m- in : *wantï ‘related by marriage, son-in-law, brother-in-law’ > Sm. vı̊ntı̊m ‘courter / bridegroom’, Nen. wennīʔ ‘related by marriage, related as brothers-in-law’, Kamass mono \ muno ‘matchmaker, suitor (acting on behalf of another)’, En. maddu ‘suitor’ In the same way, if PU *-n once existed, the same would work for : *wiδewen ‘marrow / brain’ > *wiδewe > F. yty, ydyn g. ‘bone marrow / core / power’, Es. üti, üdi g. ‘marrow’ *wiδeme > Erzya udem ‘marrow / brain / intellect’, EMr. vem, Ud. viym \ vim, Z. vem, X. welǝm, NMi. vāl(y)m ‘marrow / brain’, Hn. velő, velőt a., veleje pd.3s. ‘marrow, pith, essence’, F. luu-ydin ‘bone marrow’, ydin, ytimen g., ‘core, kernel, pith, nucleus, the central part of something, essence’, Sm. *ëδëm > NSm. aδa, aδδam- ‘marrow; marrow bone; *fat > plumpness’ It is hard to imagine another sound change that would fit either, let alone both (see other ex. below). There is no need to ignore the obvious when it requires optionality; such reliance on theory over evidence would only lead to irrationality. It is also impossible to ignore that PU *wiδewen ‘brain’ would be unrealistically close to PIE *widwon- ‘knowing / wise’ > S. vidvā́n, *wi’wön- > *w^iwwen- > TB ūwe ‘learned’, *wid-bon- > H. witpan- ‘brain(s)’ [with w-dsm., Whalen 2024a]. Many times before, I’ve said that PIE *- o:r > *-o:n > PU *-ö:y > *-ey > *-e (*wodo:r > *wöde:y > *wedey > PU *wete). If *widwo:n > *viδvẽy \ *viδmẽy, it woud show that these were caused by *-on > *-õn, etc., 1st. In Tocharian, *d optionally became *dz > ts or optional *dC > C, *dy > yy. I think this fits with *dC > *’C first, and if the same in PU, it would also fit with V > [-high] before ’ (seen in many languages throughout the world) : PIE *widwaH2- ‘wisdom / brain / intellect’, *wi’wa: > *u’iwa > *o’iwa > PU *ojwa ‘head / brain / intellect / peak / top / best’ > F. oiva ‘fine, splendid’, *oajvē > NSm. oaivi ‘head / intellect’, Mr. vuj ‘head / end / treetop’, Smd. *åjwå > Mator ajba, En. eba, Nen. ŋaywa ‘head / brain’ In the same way, *wãntï ‘related by marriage’ is close to *bhondhH-to- ‘joined / in an alliance / related / fixed’, part of other ex. of IE *bh- > PU *w- unless followed by *w \ *u (2). I have also used *CVN- > *NVN- in (Whalen 2025a) :

2025

A. Yazghulami rakW- ‘suck’ is cautiously (in “(?)” ) put together with Iranian *hrab- ‘suck / etc.’ in Cheung. Since other KW in Yazghulami are created from *vK or *Kv (or some next to *ō from earlier groups of *V(C) ), this requires Ir. *hraPK- > *hravk- or similar. Since I already have *srePH3- ‘slurp / gulp / sip’ (Whalen 2025b) to account for *-bh- vs. *-b- & problematic forms like *srpH3- > Kh. šruph- ‘slurp’, *sH3robh- > *sarobh- > H. ša-a-ra-pi /⁠ *šārabi ‘sips’, it makes sense for this to also solve the *k in Yazghulami rakW-. Ir. retained *H for a long time, often with effects on other C’s (Kümmel, Whalen 2025a), if *H > *x > *k next to *b (or only when *bH > *bhH ?; this would be undetectable in Ir. when *b(h) > *b later), it would allow *srebH3- > Ir. *hrabx- > *hrab- in most, > *hrabk- in PYz. would solve this and lend more support to *-H- in this root in PIE. B. Cheung has Ir. *taš ‘to make, construct; to cut’. This does not account for all data. In PIE *tetk^-ti > YAv. tāšti 3s., it is clear that some additional *C remained to cause *tetk^-ti > *tats^-ti > tāšti. Based on other Ir. sound changes, there is no reason for this not to be *tats^-ti > *taθs^-ti > *ta_s^-ti > tāšti. However, in OP this appears as *som+tetk^- > ha(n)taxš- ‘to work with, effect’, with apparent *tk^ > *k^s > xš. This is unexpected, but surely can not be analogy as Cheung has it. Since this is such a common root and has many odd outcomes due to *-tk^- in other IE, why would another oddity be analogical? I would expect the most common roots to cause analogy in others, not the reverse. The root he assumed caused it, *thwaxš, is not really likely to have affected *taš (if it really existed in this form, or with Cheung’s meaning in OP). I have also (Whalen 2025c) tried to use alternation of IE ts / ks to explain other apparent *k^ > (x)š before C : > If PIE *k^ was pronounced *kx^ / *ks^ / *ts^ / *tθ^ at one stage in PIr., this could have created *k^t > *kx^t > xšt in : *prek^- > L. prēx ‘request’, *prek^-tor- > Av. paiti-fraxštar- ‘interrogator’ *spek^- ‘look at’ > Av. spaxšti- ‘vision’, spašta, S. spaṣṭá- ‘clearly perceived/discerned/visible’, L spectus, speciō *y(e)H1-k^-?? > L. iacere ‘throw’, *ya(x)śt- > Av. yaxšti- ‘branch’, S. yaṣṭí- ‘stick/staff’ That each group of problems is optional seems clear, so attempting to find more regularity than the data provides would only be counterproductive. Without full knowledge of PIr. dialects, idiolects, free variation, etc., pretending that current irregularity can not come from an older stage with regularity, or even that it can not be real, is not warranted. > There is no added problem by including this data. Since many other linguists (Alexander Lubotsky, Jay Jasanoff) have seen simply *k^t > (x)št with no additional descriptions of intermediate forms, I wonder why Cheung would need analogy here. If correct, it would only mean that all other *k^(C) > xš would need analogy, or at least not be from a sound change to *k^(C) itself. To complicate things, other IE cognates show changes that can’t be regular (if all recontructions are right). Ct. *tetk^-(t)lo- > *tetsklo- > *taktslo- > *taxtslo- > *ta:tslo-, *tetk^on- ‘carver / maker’ > *tetskon- > *tektson- > Gmc *þixtsan- > ON Þjazi (based on his appearance in a myth similar to one with Tvaṣṭṛ < *twerk^- ‘cut / carve’ (2025d)). Due to problems like this, I see no way for all the outcome of *TK to be regular (2025e). It is theoretically possible that these clusters were actually *TTK or *TKK or some other odd form that might provide regular outcomes. Even velar vs. uvular might work. However, due to many other IE sound changes that seem to be optional, I see no real need for this. Some of these show early *VTK > *V:K or *VtK > *VtsK before later *TK > KT \ KS in most, so a stage in which tk \ tsk \ ’k were in free variation would be regular at that time, but when only one variant lasted, it would appear to create irregularity “from nowhere”.

2025

Summary: Clayton analyzes many *r > ur vs. ir in S., some based on rounded CW. This includes more than traditional PIE *kW, etc. Khoshsirat & Byrd (2018, 2023) argue that the IIr. causatives (S. -āpaya- & Ir. *-āwaya-) come from *oH > *oHW (or similar). I have additional evidence of new CW in IIr., created by w / u / o, as well as examples of *H > f and *s > f in IE. This includes *H3 (likely = xW / RW) becoming w or f; *o or *u causing rounding to fricatives (Al. *H > p, G. *HW & *hW (from *s) > h / 0), IIr. *-os > *-av, and many others. This is part of a broad change of *H & *s being rounded by nearby (not always adjacent) round or labial C / V. In its nature & optionality, it also resembles n > m, s > f, etc., in the same environments

2025

There are many cases of optional *p > k near P / w / u in S., sometimes also in Iranian : *pleumon- or *pneumon- ‘floating bladder / (air-filled) sack’ > G. pleúmōn, S. klóman- ‘lung’ *pk^u-went- > Av. fšūmant- ‘having cattle’, S. *pś- > *kś- > kṣumánt- \ paśumánt- ‘wealthy’ *pk^u-paH2- > *kś- > Sg. xšupān, NP šubān ‘shepherd’ *pstuHy- ‘spit’ > Al. pshtyj, G. ptū́ō, *pstiHw- > *kstiHw- > S. kṣīvati \ ṣṭhīvati ‘spits’ *pusuma- > *pusma- > S. púṣpa-m ‘flower/blossom’, kusuma-m ‘flower/blossom’ *tep- ‘hot’, *tepmo- > *tēmo- > W. twym, OC toim ‘hot’, *tepmon- > S. takmán- ‘fever’ *dH2abh- ‘bury’, *dH2abh-mo- ‘grave’ > *dabH-ma- > *daf-ma- > YAv. daxma- S. nicumpuṇá-s \ nicuṅkuṇa-s \ nicaṅkuṇa-s ‘gush / flood / sinking / submergence?’, Kum. copṇo 'to dip’, Np. copnu 'to pierce, sink in’, copalnu 'to dive into, penetrate’, Be. cop 'blow', copsā 'letting water sink in’, Gj. cupvũ 'to be thrust’, copvũ 'to pierce' S. kṣubh- ‘shake’, Pa. chubh- ‘throw out’, *tsup- > L. supāre ‘to throw/scatter’, Li. supù ‘I rock (a child in a cradle)’, *tsok- > *kṣot- > S. kṣoṭayati ‘throws’ [retro.-asm.] I proposed (Whalen 2025a) that these were cases of *P > *KW near P / w / u. This is based on changes in Iranian that looked like *KW > P near *KW in :

Journal of the Phonetic Society of Japan Vol. 28: pp. 52–60, 2024

Of the two Tocharian languages, i.e. Tocharian A and B, it is well known that the Proto-Tocharian cluster *-kS- is sometimes reflected as -pS- in Tocharian A whereas the same cluster is preserved as -kS- in Tocharian B. While there are a number of brief references to this phenomenon in the literature, no comprehensive investigation of the available data and nature of this change has been carried out. Here I collect all of the known examples and discuss the conditions under which this change may occur and the nature of the change from a crosslinguistic perspective. I argue that *-kS- > -pS- can be understood as an instance of CHANGE within the framework of Blevins' (2004) CCC-model of sound change.

Mashoori is a dialectal variation of southeast of Khuzestan Province, Iran. This dialect is spoken in Mahshahr and its regional neighborhood and, inherited from the Old Southwest Iranian Languages. It is somehow rich from the phonological perspective. One interesting phonological process applied in MD is the alteration of consonantal /b/ and /f/ to the semivowel [w] that is assumed to be kind of lenition process. /w/ in the Old Southwest Persian changed historically to [g], [b] and [w] in Persian Dari and to [g], [b] and [v] in Modern Standard Farsi. The data required for this descriptive-analysis study make use of a dialectal report manual gathered by five native speakers of MD, plus data collected from previously done related researches on this process within the theoretical framework of Generative Phonology. Findings render that this process is remarkably different from what is going on in Modern Persian in a way that seems plausible to choose /w/ as the underlying representation, changing to /b/, /f/ and /v/ in Modern Standard Persian under the fortition process.

2025

Indo-European Roots Reconsidered 22: *H2aws-r, *H2wes-r, *wesH2-r ‘spring’ Sean Whalen stlatos@yahoo.com April 21, 2025 There are disputes about whether PIE ‘spring’ & ‘dawn’ are related. I think evidence of several types of laryngeal metathesis in cognates (Whalen 2025a) makes their relation clear. Looking at S. vasar ‘at dawn’, Av. vaŋri ‘in spring’; S. vāsará- ‘relating to morning’, OP Θūra-vāhara- ‘(month of) spring swelling/growing’ it seems impossible to separate them in a reasonable way. A retention of the older meaning in S. makes much more sense than metathesis of *awsar within S. happening to create 2 words that looked identical to ‘spring’, both happening to refer to early time periods. The shift ‘early part of day’ > ‘early part of the year’ makes an origin from a verb indicating time likely (Whalen 2025a), with *H2wes- ‘stay (the night) / (stay until) dawn’ the only good choice. Looking at IE cognates, a huge number of irregular changes and many types of metathesis are needed, showing that optionality was common in IE : *H2aws-r, *H2wes-r, *wesH2-r, *ewsH2-r ‘spring’, obl. *-n- *ewsH2-r > TA yusār ‘rainy season?’ (Pan) *H2ant-wesH2n- ‘early spring’ > H. hamešha(nt)- \ hameškant- ‘spring / early part of the year’ [n-n > m-n, mtw > mw no other ex.] *H2wesr > S. vasar-hán- ‘destroying (nocturnal demons) at dawn’, Av. vaŋri l. ‘in spring’, MP wahār, [irr. *(t)sr, Kümmel] Zz. wesar, Tal. ǝvǝsor, G. éar, Ion. êr, Hsx. géar = *wéar nu., earīnós aj., *werǝr > *werr ? > L. vēr nu., vē̆rnus aj., U. Urnasier p.d/abl. ‘an early spring month’, Gmc *wezr- > *wǣra- > ON vár (Gąsiorowski) *H2wesn- > OCS vesna ‘spring’ *H2wesr-ako- > *xWexrako- > *xexrako-? > OI errach ‘spring’ *H2wesr-onto- > Ar. garun, garnan g. [not **gaṙnan, indicating old *garǝnan < *garǝndan; n(d) < *nt in other words, not reg.] *H2wes(n)-onto- > S. vasantá- m. ‘spring’, Pl. basaán(d) m., basandá p., Ks. básond \ básund, Kh. bosùn, Sh. bʌzṓno, Ti. bǝsãn, Kv. vâsút, *va:sút-vór > vâsdór ‘summer’, Sa. vâsanta ‘summer’ Ct. *wehant-eino- aj. > OW guiannuin, MW gwaeanhwyn, W. gwanwyn, OCo. guaintoin Ct. *wesn-aHl\alH-aH2-? > MW gwennawl, [e-a > a-e] OI fannall f., fainle g. ‘swallow’ S. vāsará- aj. ‘relating to morning’, m/nu. ‘day’, OP Θūra-vāhara- ‘(month of) spring swelling/growing’ *H2awsr > *H2wasr > Gmc *warsa- > OFr wars ‘spring’, Li. vãsara \ vasarà ‘summer’, vasarìnis aj. *H2awsr -> Gmc *austra- \ *austro:n- > OHG Óstara, OE Éaster \ Éastre, E. Easter Pan’s *isu- ‘foaming -> *yus-ar > TA yusār ‘rainy season?’ does not seem needed, and the metathesis in so many other cognates shows that *we- > *ew- fits the context. Though *-H2r > *-ar is possible (also *H1esH2r > *yäsar), most other PIE *-r > PT *-är > *-ar, maybe regular (Whalen 2024a), and with 4 ex. it would be pointless to say all of them came from “collective *-o:r” unseen in any cognates : *H1itr > *yitär > *yätär > *yätar > TA ytār, *-yo- > TB ytārye ‘road / way’ *H1esH2r > *yesär > *yäsär > *yäsar > TB yasar ‘blood’ *g^hesr > *kesär > *kyäsär > *k^äsar > TA tsar, TB ṣar ‘hand’ If 1st ‘early part of the year’, the compound *H2ant-wesH2n- with *H2ant- ‘in front / before / early’ makes sense for H. hamešha(nt)-. Though Kloekhorst said *ntw > w would not be reg., there is no way to know what *mtw might become after *n-n > m-n, part of many IE alternations of m / n near n / m & P / KW / w / u (Whalen 2025b), and even *tw-t > *w-t is possible in forms with -ant-. For *sx > šh \ šk in hameškant-, Kloekhorst said it was irrelevant, but see Weiss for other ex. and cause of h \ k. MP wahār supposedly had analogy with *vāhara- (OP +vāhara-) & metathesis of length. Since *H2wesr contained *H, early H-metathesis seems more likely than unmotivated metathesis of a feature to an unexpected place, and H-metathesis was very common in Ir. (Whalen 2025d), seen by devoicing C’s. In MP wahār vs. Zz. wesar, irr. *(t)sr in Ir. (Kümmel, Whalen 2025c). Other cases of *sr > *tsr > θr in Ir. include : S. sraktí- ‘prong/spike/point / corner/edge’, Av. sraxti- \ θraxti- ‘corner’ S. srotas-, OP rauta, Av. θraōtah- ‘river’, raōðah- ‘stream’ *tem(H)sro- ‘dark’ > S. támisra-, tamsrá-, Av. tąθra-, Li. timsras Gmc *wezr- > *wēr- > *wǣra- > ON vár comes from stress in the obl. cases, generalized in most, with *zr changed as in Gąsiorowski. For *H2wesr-ako- > *xWexrako- > *xexrako-? > OI errach ‘spring’, I doubt that expected *ferrach was lost by analogy after V. Though both *f- > 0- & *0- > f- are fairly common later, here the old attestation might be best solved by asm. of *xW-w after *w- > *xW-, before *xW- > f- (if this timing works). In my *H2awsr > *H2wasr, since there is no other ev. for *wosr with o-grade, another case of laryngeal metathesis is best, since metathesis is needed for words in which different e- vs. o-grades would solve nothing. Adams, Douglas Q. (1999) A Dictionary of Tocharian B http://ieed.ullet.net/tochB.html Baart, Joan (1997) The sounds and tones of Kalam Kohistani: with wordlist and texts https://www.academia.edu/1992270 Baart, Joan (2005) A first look at the language of Kundal Shahi in Azad Kashmir https://www.academia.edu/1992366 Bashir, Elena (1988) Topics in Kalasha syntax: an areal and typological perspective https://www.academia.edu/82507617 de Vaan, Michiel (2008) Etymological Dictionary of Latin and the other Italic Languages (Leiden Indo-European Etymological Dictionary Series; 7) Decker, Kendall D. (1992, 2004) Sociolinguistic Survey Of Northern Pakistan Volume 5 Languages Of Chitral Gąsiorowski, Piotr (2012) The Germanic reflexes of PIE *-sr-in the context of Verner's Law https://www.academia.edu/64951212 Kloekhorst, Alwin (2008) Etymological Dictionary of the Hittite Inherited Lexicon https://www.academia.edu/345121 Kümmel, Martin Joachim (2012) The Iranian reflexes of Proto-Iranian *ns https://www.academia.edu/2271393 Liljegren, Henrik (2009) The Dangari tongue of Choke and Machoke: Tracing the proto-language of Shina enclaves in the Hindu Kush https://www.academia.edu/3849218 Liljegren, Henrik (2010) Palula vocabulary https://www.academia.edu/3849251 Liljegren, Henrik (2013) Notes on Kalkoti: A Shina Language with Strong Kohistani Influences https://www.academia.edu/4066464 Lunsford, Wayne A. (2001) An Overview of Linguistic Structures in Torwali, A Language of Northern Pakistan https://www.fli-online.org/documents/languages/torwali/wayne_lunsford_thesis.pdf Martirosyan, Hrach (2009) Etymological Dictionary of the Armenian Inherited Lexicon https://www.academia.edu/46614724 Matasović, Ranko (2009) Etymological Dictionary of Proto-Celtic https://www.academia.edu/112902373 Pan, Tao (2024) Notes on the Tocharian A Lexicon https://www.academia.edu/128459731 Perder, Emil (2013) A Grammatical Description of Dameli Rajapurohit, B. B. (2012) Grammar of Shina Language And Vocabulary (Based on the dialect spoken around Dras) Strand, Richard (? > 2008) Richard Strand's Nuristân Site: Lexicons of Kâmviri, Khowar, and other Hindu-Kush Languages https://nuristan.info/lngFrameL.html Turner, R. L. (Ralph Lilley), Sir. A comparative dictionary of Indo-Aryan languages. London: Oxford University Press, 1962-1966. Includes three supplements, published 1969-1985. https://dsal.uchicago.edu/dictionaries/soas/ Weiss, Michael (2016) The Proto-Indo-European Laryngeals and the Name of Cilicia in the Iron Age https://www.academia.edu/28412793 Whalen, Sean (2024a) Notes on Tocharian Words, Loans, Shared Features, and Odd Sound Changes (Draft) https://www.academia.edu/119100207 Whalen, Sean (2025a) Indo-European Roots Reconsidered 21: *H2aws-, *H2wes- ‘(stay until) dawn’ https://www.academia.edu/128907134 Whalen, Sean (2025b) IE Alternation of m / n near n / m & P / KW / w / u (Draft 3) https://www.academia.edu/127864944 Whalen, Sean (2025c) Indo-European Roots Reconsidered 4: Sanskrit pāṃsú- / pāṃśú-, síkatā- https://www.academia.edu/127260852 Whalen, Sean (2025d) Laryngeals and Metathesis in Greek as a Part of Widespread Indo-European Changes (Draft 6) https://www.academia.edu/127283240 https://en.wiktionary.org/wiki/Reconstruction:Proto-Germanic/wazr%C4%85 https://en.wiktionary.org/wiki/Reconstruction:Proto-Italic/wezor

2025

A. It is known that Greek words show alternation of l \ d, from either *l or *d : G. dik- ‘throw’, dískos, Perg. lískos ‘discus / disk / dish’ G. Odusseús \ Olutteus \ Ōlixēs < *wlkWo- ‘wolf’ or *luk- ‘bright’ G. *Poluleúkēs ‘very bright’ > Poludeúkēs ‘Pollux’ (like Sanskrit Purūrávas- ‘*very hot’) G. dáphnē / láphnē, NG Tsak. (l)afría, L. laurus ‘laurel’ LB ko-du-bi-je < *kolumbiyei (woman’s? name) << *kolumb- ‘dove’ (6) LB da-bi-to ‘place (name)’ < *Labinthos, G. Lébinthos *molHo- > L. mola ‘millstone / grains of spelt (& salt)’, G. môda ‘barley meal’ *polo-s > G. psólos ‘soot/smoke’, spodós ‘(wood-)ashes/ember/dust/oxide/lava’, spódios ‘ash-colored’, spoleús ‘loaf of bread’ G. kélados ‘noise/clamor / sound/cry/shout / twitter/chirp’, *kelalúzō > kelarúzō ‘murmur’ G. kálathos ‘basket with narrow base / cooler (for wine), *kadath-? > Arc. káthidos ‘water-jug’ *laHk-? > L. lacerna ‘a kind of cloak, worn over the toga’, *lVkk-? > G. lákkos ‘a kind of garment’, lókkē ‘short mantle’, lékkē \ dektḗ ‘upper-garment / cloak / wrapper, worn loose over the chiton’ but some also include th : G. alṓpēx ‘fox’, Pontic G. thṓpekas \ thépekas >> Ar. t’epek, MAr. t’ep’ēk \ t’obek ‘jackal’ G. dáptō ‘devour/rend/tear’, dáptēs ‘eater / bloodsucker (of gnats)’, Cr. thápta, Pol. látta ‘fly’ (Witczak 1995) which would fit if *ð > l and *l > ð were due to varieties of G. having *d & *th as fricatives much earlier than others. In other IE, ð > l is fairly common (Iranian). Some of these have been seen as loans from Anatolian languages (some of which had *T > l, though also not apparently regular), but if other IE branches had alternation of l \ d, this would be much less likely. B. Italic words also show alternation of l \ d. Latin irregularly changed both *d and *dh to l. Examples of *d > l : *H3od- ‘smell, stink, hate’ > L. oleō, odor, odī ‘I hate’, Ar. hot *mazdo- > I. maide ‘stick/staff’, L. mālus ‘mast’ *dH2ak^ru- > OL dacruma, L. lacrima, G. dákru \ dákrūma, Go. tagr *sodiyo- > OI. suide, Gaelic suidhe ‘seat, sitting’, L. solium ‘seat, throne’ *smeru- > OE smeoru ‘fat/grease’, NHG schmirwen, E. smear, OI smiur, TB ṣmare ‘oil/*oily>smooth’, L. melilla \ medulla ‘marrow’ and many more. Also *dh > l in *mizdho- > G. misthós ‘wages’, L. mīles ‘soldier’. It is likely this also shows *d(h) > *ð before *ð > l, especially because it’s very common in *zd(h), implying a change due to fricative-assimilation *zd > *zð. If there was optional metathesis of aspiration in *dng^hwaH2- > E. tongue, *dhng^waH2- > L. dingua > *ð- > lingua, *thǝŋgwa: > Umbrian fangva- (impied by U. *th- > f-), then this would be part of *dh > l instead of standard *d > l (optional *dh-g > l-g vs. *d-gh > d-g). These changes are just as clear now as they were then, whether due to Sabine influence or not. Prósper said, “the Sabine attribution [of d > l] is a modern myth, never explicitly found in the writings of the ancients.” Whatever the source, they are no more regular than in Greek. Prósper said that *da(:)- > la(:)- was regular, but many of the best examples are not before -a-. This seems old enough to be due to *d > *ð near *H, and, of course, most *a were caused by *H2 or syllabic *H. Though most linguists say all these words are from *d, some are of unknown origin. Since James Clackson argues for *kl > *kð in South Picene (kduíú (L. clueō), brímeqlúí and *brēmekdīno- > brímeidinais, qdufenio- >> L. Clufennius), it makes sense for Italic, or some sub-group if all Latin l \ d is due to foreign influence, to have optional alternation of l \ d like Greek. Indeed, some of these words are G. loans, in which the timing might allow l \ d in either language : G. thṓrāx, Ion. thṓrēx ‘corslet / coat of mail’, L. lōrīca ‘coat of mail / breastplate’ G. númphē, L. lumpa ‘nymph, (spring) water’, Oscan *dümpa > diumpa- (with dissimilation of nasals n-m > l-m) I think optional *kl > *kð in SPc and *d > l in Latin is part of this broad change of *ð > l and *l > *ð > d. Prósper also described L. d > Romance l as due to a stage with d > *ð. Depending on timing, *th > l might also exist. In L. ūvor ‘moisture’, ūvor ‘liquid/fluid/moisture’, ūlīgō ‘moisture’, ūmidus \ ūvidus \ ūdus ‘moist’, ūmēre ‘be moist’, ūmēscere ‘become moist’, several alternations seem to exist, but if statives in *-eH1- had aj. in *-H1to- > It. *-atho- (Rasmussen 2007, Whalen 2023a, 2024c), then it is possible that *u:wathos > ūdus, *ūd- > ūlīgō. Since *kl > *kð > kd is one of the last places you might expect to see *l > *d, is there other evidence for *Cl > *CT ? I think that *slaH2no- > OI slán, slántu ‘health’, L. sānus ‘well/healthy’ might show *sl- > *sθ- > s-. If not, sānus would be isolated, and its resemblance to slán is often noted. Any Italo-Celtic theory should examine such potential cognates carefully. C. Tocharian also seems to have *th > l (Whalen 2025a) : > Bb. G. has many -thmo-: porthmos ‘ferry/strait’, iauthmós ‘sleeping place (of wild beasts)/den/lair’, arithmós ‘number’. It is likely this corresponds to L. -timus < *-tmHo- with H-met. (Whalen 2025c) causing aspiration: *-tmHo- > *-tHmo- > -thmo-. This also has to do with a solution to Tocharian -lme. If from IE, what created *-lmos? Since Toch. shared features with Greek (like breaking related to H123, H1 > i, etc.), why not this too? It would show likely *th > l (common in many, including G. dáptēs ‘eater / bloodsucker (of gnats)’, Cretan thápta, Polyrrhenian látta ‘fly’; with each stage shown by the alternation). Both PT and G. would have the odd changes to *-tmHo- and some *th > l (likely dia. in G., maybe reg. in PT). Together, PT *-θmos > *-θme > -lme, acc. *-θmom > *-lm’äm > *-ln’äm > [ana.] > *-ln’e(m) > -lñe. An interdental stage would unite changes to PT *th and *s in a common stage. If *s > *θ adjacent to *s, *CsC > *sC, *θs > ts, *θ > l : *H2wes- > OE wesan ‘be/remain’, S. vásati ‘dwell’, G. aes- ‘spend the night / pasture’ *H2wes-sk^e-, G. aéskō ‘*spend the night’ > ‘sleep’, *wäθsk- > *wäθk- > *wälk- > TB woloktär ‘dwells’ (with Csk > Ck (as in many -tk- verbs) and the same developments as *kWelH1- > koloktär ‘follows’ ) > Since I said that *-om caused several TB alternations (Whalen 2025a), I consider the suffixes TB -(e)lñe & -(e)lme related, with *m-m > *ñ-m. Just as *-to-s & *-to-m ( > *-tem ) > -te & -ce, also *-thmo-s & *-thmo-m > -lme & -lñe. Adams also considered a “special phonetic development of of pre-Tocharian *-δn- in a nasal present” (1) : *lH1d-ne- > *lədne- > Al. lë ‘let’, *laðne- > *lalnä- > TB lāl- ‘exert oneself / strive for’, cau. ‘tire / subjugate’ Of course, a “special phonetic development” is simply an irregular change, however worded. It did not happen in his *moudno- > TB maune ‘avarice’, for example. Indeed, it need not be any more regular than *d(h) > d / l in some Latin words. He also had TB luwo ‘animal’ from OCS loviti ‘hunt’, SC lov ‘game animal’, but with “a cross of this etymon with a PTch *tsuwā ‘animal” (2). Why is such a “cross” needed if he already had some *-dn- > *-ln-, also not regular? Why include ‘game animal’ when ‘animal’ has all the features needed, and no *-s- in *lewo-, etc.? This would just be : *dhewHso-m > Go. dius ‘wild animal’, OE déor ‘animal’, E. deer, Li daũsos f.p. ‘upper air’, Sl. *dûxŭ ‘gust/breath/soul/spirit’, OCS duxъ ‘spirit’, OR duxŭ ‘air’ *dhewHos-s > *lewo:s > TA lu, lwāk p., TB luwo, pl. lwāsa ‘animal’ *dhewHos- > *dhowHos- > *dhwoHos- -> en-dhwoHos-s ‘having spirit inside’ > H. antuwahas n., antuhsan a. ‘(hu)man’ >> H. antūh Since *d > *dz > ts is usual (whether before front or back V), but some *d > t, the stages *d > *d / *dð, *dðe > *dðiä > *dð’ä > *dðä, *ð > *z \ *l can explain why not *lyuwo. In support, there are other ex. of *d > l, also optional. Since TB also had some r \ l (Whalen 2025b), I also see *d > l related to several examples of *nd > *nr. It is possible that when *d > *ð > l, *nð > *nr if these examples are complete & relevant, but it is possible that others have been missed or are unattested. Ex. : *dhewHos-s > *lewo:s > TA lu, lwāk p., TB luwo, pl. lwāsa ‘animal’ *leH1d- > G. lēd- ‘be tired’, Al. lodh tr. ‘tire’, *lH1d-to- > L. lassus ‘weary’ *lH1d-ne- > *lədne- > Al. lë ‘let’, *laðne- > *lalnä- > TB lāl- ‘exert oneself / strive for’, cau. ‘tire / subjugate’ S. saṁdhí- ‘junction, connection, combination, union with (+ instrumental) / association, intercourse with (+ instrumental) / comprehension, totality / agreement, compact / alliance, league, reconciliation’ >> PT *sanri > TB sārri ‘assembly’ *en-diwyos > G. éndīos ‘in the middle of the day’, *iänduwos > *enduwe > *endwe > *enrwe > *nerwe > TB ñerwe ‘today’ *H3ozdo- ‘branch’ > Ar. ost, G. óz[d]os, Go. asts, *oz(ä)do- > *esäle > TA asäl, TB esale ‘post’ *pezd- > L. pēdis ‘louse’, pazdu- ‘maggot’, *pozdo- > TB peṣte ‘worm? / maggot? / louse?’, peṣteu ‘worm/lice-ridden?’, peṣele ‘some kind of unlikeable insect’ (3) TB yälloñ < *Hed-lo- or *wid-lo- (since most *dC > C, *dl > ll would show a special outcome, fitting other unconditioned *d > l) (4) maybe also : S. padá-krama- \ krama-pada- ‘series of steps / pace / series of quarters of verses / ~method of reciting the Veda’, In. *krama-padyā >> PT *krämä-pädyā > *krämä-pälyā > TA klumpri, TB klampärya ‘~meter (4x18 syl., rhythm 7/7/4)’ (or *d > r, *r-r > *l-r later) *se-s(e)d- > *sezd- > G. héz[d]omai ‘seat oneself / sit’, Av. hazdyāt op.3s *sezd-ne- > Ar. hecanim \ hecnum \ hejnum ‘mount a horse / ride’, *siäzðmä- > *syäzmä- > *šämä- > TB ṣäm- ‘sit’, *syäzlmä- > *slyäzmä- > lyämā- pt., lyämäsk- cau. ‘set’, etc. This is not related to *lemb- > E. limp, S. lamb- ‘hang down’, TB läm- ‘cling to’ (5). In *sezdne-, needed for Ar., a change of *s-sCn > *sC-sn or similar would not be odd, so *T > l here can not be timed exactly. Clearly, there are enough cases of *T > l for this to be as clear as in Latin, even if not all are as certain as others. Since also...

2025

Witczak had, in part, *wrk^- ‘be fat’ ? > H. warkant- ‘fat’,*wrk^- > G. Cr. árkos / árkālos / arkḗla ‘badger’, NG Cr. árkalos, T. *wVrk(V) > KxM wark, *w(o)rk^- > Ar. goršuk, Np. bharsia ‘(honey) badger’, NP barsū(kh), Kd. barsuk; many also say Tc. *bors(m)uk was an Ir. loan. The reasons for including KxM wark is that it was an animal similar to Tc. borsmuk, seen as a symbol of fatness. The T. loan is a good idea due to geography & history. I agree with his basic ideas, but more can be said. Ar. goršuk requires *work^wuko- (or *work^wu:ko-, etc.), with the same *k^w > *s^w > *s^y > š in *k^uwo:n > *k^wu:n > *syun > šun ‘dog’, *H1ek^wo- ‘horse’ > *ešyo > *eyšo > ēš ‘donkey / ass’. Clearly, most IE would have *Cwu > Cu. However, other evidence of *-k^wu- exists here. Tc. *bors(m)uk was clearly from *worswuk, with optional dsm. *w-w > *w-m before *w > *v > *b. If no dsm. in any variant, *Cwu > Cu like normal. This explains *worswuk > *borsuk vs. *worsmuk > *borsmuk, in which -m- appearing “from nowhere” is not just something that can be passed over in silence (yet it has previously). The -o- corresponding to Ar. -o- also can’t be found in Ir. It would be impossible if really an Ir. loan from something like barsuk, so why is this theory so prominent? It is only needed if all similarities between Tc. & IE need to be loans, however much they might not fit. Since borsuq vs. barsuk can’t just be waved away, it must be a loan, in their view. This mix of features requires some IE language now unknown, but not TA, TB, since T. *wVrk(V) > KxM wark shows a separate form. There are many Tc. >> Tocharian loans, but those said to be Tocharian >> Tc. loans by Ünal are very odd, and show some changes not expected due to timing, etc. (Whalen 2025a). I also find it impossible to believe PT was so prominent that it could influence PTc. so much. This would require, at least, an IE language similar to PT that was ancestral to PTc. Since *work^wuko- is such an odd form, it requires more analysis. The PIE word *work^-wo:s ‘having fattened (oneself) / grown fat’ would work, since in other cases *-wot-, *-ut-, *-us- existed (or strong *-wos- vs. weak *-ut- were mixed in later IE). Whatever the original, some IE turned *-us- > *-wus-, etc., by analogy with *-wos-. This allows *work^-wo:s -> diminutive *work^-wut-ko-s. Some IE *VdK > *V:K, and I have said that outcomes of *TK were not regular (Whalen 2024a, 2025b), and I take this as further evidence. For ex., in *VdK > *V:K, if it were of PIE date, how would *VdKn > *VnKn in IIr.? Other words clearly show *tK > *tsK or *K(t)s, with no regularity. In *work^wutkos > *work^wu:kos, the odd *-o-u-o- is explained by the same change as *VdK > *V:K. In all : *wrk^- > G. *wárkos > Cr. árkos / árkālos / arkḗla ‘badger’, NG Cr. árkalos, T. *wVrk(V) > KxM wark *work^-wo:s ‘having fattened (oneself) / grown fat’ *work^-wut-ko- > *work^wu:kos > Ar. goršuk, Np. bharsia ‘(honey) badger’, NP barsū(kh), Kd. barsuk *wörswu:kö > *bors(m)uk(ï) > OUy bors(m)uk, Kx. bors(m)uq, Ui. borsuq, Tk. porsuk, Khk. p\morsïx, Tv. morzuq, ? >> Hn. borz

Minos, 2016

This article studies a remarkably consistent peculiarity observed in the orthography of the attested oblique cases of <wa-na-ka> ϝάναξ in Linear B (Genitive <wa-na-ka-to> ϝάνακτος and Dative <wa-na-ka-te> *ϝανάκτει), as well as its derivative adjective <wa-na-ka-te-ro> *ϝανάκτερος. In these types, occurring in documents of wide geographic distribution and by a considerable number of ‘scribes’, the clusters /-kto-/ and /-kte-/ are represented arbitrarily as <-ka-to> and <-ka-te->, with a peculiar ‘obstinate’ use of an Orthographic (Arbitrary) Vowel <-a-> instead of the more ‘canonical’ Orthographic Copy Vowel that adopted the value of the succeeding phonological vowel (<-ko-to> and <-ke-te-> respectively). The analysis presented here suggests that the <-a> vowel in the Nominative Singular type <wa-na-ka> ϝάναξ is an Orthographic Arbitrary Vowel, retained, against ‘canonical’ practice in the other ‘idiosyncratic’ <wa-na-ka-> spellings. Taking into account the lack of any convincing IE etymology for ϝάναξ and its most probable status as an adopted loanword, we have proposed to associate this with other spelling ‘idiosyncrasies’ that occur in grapheme-sequences that were arguably ‘transmitted’ from Linear A into Linear B, such as <su-ki-ri-ta> or <pa-i-to>. In this way, we have argued for that the <wa-na-ka-> spellings may indicate a ‘graphemic reception’ of the loanword (resulting in its orthographic fossilization), as opposed to a phonetic adaptation. If so, the source of such a ‘graphemic borrowing’ could only have been the ‘parent’ system on which the Linear B syllabary would have been basically modelled which seems to have been a variant of the Neopalatial Linear A script. This would positively suggest that <wa-na-ka> was adopted into Greek from a ‘Minoan’ source. The key significance of the term (the title of a ruler), as well as the uniquely broad distribution and consistency of the <wa-na-ka-> spellings (suggestive that these spellings formed part of the ‘core’ tradition in Linear B scribal training) enable us to associate safely its adoption with the complex set of events surrounding the genesis of the Linear B writing and administrative system, which most probably took place in Late Minoan IB or II at Knossos. This line of argument would positively suggest the ‘Minoan’ origin of the title -a possibility anticipated by previous scholarship- without hazarding hypotheses regarding the identification of the ‘Minoan’ language(s).