LINGUIST List 21.2719

Fri Jun 25 2010

Sum: /f/ to /theta/ Sound Changes

Editor for this issue: Danielle St. Jean <daniellelinguistlist.org>

        1.    Ryan Bennett, /f/ to /theta/ Sound Changes
Message 1: /f/ to /theta/ Sound Changes
Date: 25-Jun-2010
From: Ryan Bennett <rbennettucsc.edu>
Subject: /f/ to /theta/ Sound Changes
E-mail this message to a friend

Query for this summary posted in LINGUIST Issue: 21.920 Some months ago I posted a query on The LINGUIST List asking forany attested instances of a historical /f/ > /theta/ change, or for anydialect variation between /f/ and /theta/ in which /f/ has become /theta/in the innovative dialect.

Many thanks to the following people for responding to my query:

James BrookesRonald CosperInge GeneeLauren Hall-LewPatrick HoneyboneMark JonesMartin KümmelKelly Lynne MaynardDaniela MüllerPamela MunroBoyan NikolaevEmerson OdangoMikael ParkvallJoseph SalmonsGary TaylorRémy Viredaz

I also have responded to each of these people individually. I apologizeto them (and all other interested folks) for not posting this summarysooner.

The responses were of four types, and are grouped below.----------(I) Attested /f/ > /theta/ changes:

Mikael Parkvall and Emerson Odango noted that in Pulo-Annan(Chuukic), Proto-Micronesian *f became /ð/ (orthographic ), despitebecoming /f/ in all other Chuukic languages.

The motivation for this change is unclear, though it may have beeninfluenced by areal contact with Palauan, which has /ð/. The Chuukiclanguage Ulithian also has /ð/, but as a regular reflex of Proto-Micronesian *d.

Joseph Salmons and Gary Taylor also pointed out an example of aconditioned /f/ > /theta/ change: Gothic reflexes of early Germanic /fl-/often show up as /theta+l/.

References:(a) Schiko Oda’s thesis “The Syntax of Pulo Annan”

(b) pp. 4-5 of Bender, Byron, Ward Goodenough, Frederick Jackson,Jeffrey Marck, Kenneth Rehg, Ho-min Sohn, Stephen Trussel & JudithWang (2003): Proto-Micronesian Reconstructions 1. OceanicLinguistics 42 (1), pp 1-110.

(c) Joseph Salmons and Gregory Iverson (1993) "Gothic /þl-/ ~ /fl-/Variation as Lexical Diffusion." Diachronica 10.87-96.

(d) Jones, Mark J. (2002). "More on the instability of interdentalfricatives." Word 53(1): 1-8.----------(II) Reconstructed (but unattested) /f/ > /theta/ changes:

James Brookes, citing Stuart-Smith (2004), suggested that Proto-Indo-European /dh/ may have passed through /f/ on its way to becomingLatin /d/. I repeat the argument more or less verbatim:

"...in Latin and Sabellian the PIE so-called voiced aspirated stop /dh/went through a lenition stage in medial positions, /dh/ > /d/ > /theta/ >/f/. In Sabellian, the lenition remained, so /f/ is found in words likemefiaí '(in the) middle' < PIE *medh-yos.

In Latin, /f/ is not the actual medial reflex of PIE /dh/, but rather /d/, sothe lexical equivalent of Sab. mefiaí is Lat. medius. Smith thereforesupposes that in Latin /f/ underwent some kind of re-fortition to /d/ inthe medial position; the only way to account for fortition is through astepwise reversal, which should naturally involve a /theta/ stage."

References:(a) Jane Stuart-Smith (2004). "Phonetics and philology: sound changein Italic." Oxford University Press.----------(III) Lexically isolated /f/ > /theta/ changes

In his own extensive diachronic work, Martin Kümmel found only onepotential case of a /v/ > /ð/ change, localized to the Eastern IranianPamir languages of the Shughni group (including Yazghulami) andneighbouring Sanglichi. This was a conditioned change, with "voiced /v/[becoming] the corresponding dental fricative before /m/ from /n/" in asingle form.

The word for 'sleep,' which "originally must have had /fn/ voiced to/vn/," ultimately reaching /ðm/, with "the labial feature...gone over to thenasal." The diachronic trajectory of this change might then have been/fn/ > /vn/ (> /vm/) > /ðm/. Martin speculated that this final stage mayhave been dissimilatory in nature.

Martin also notes that (unsurprisingly) /theta/ > /f/ changes werewidespread in his survey (p.193f).

Daniela Müller found a single case of /f/ > /th/ in a dialect of Occitan:Latin /febrem/ 'fever' is realized as /ther/. However, this may notrepresent a true /f/ > /th/ change, but rather a mediated /f/ > ... > /fth/ >/th/ change. As Daniela writes:

"...it is an /fj/-cluster that developed into the interdental fricative,because of the diphthongisation of the Latin short /e/. The palatal glidein stop+palatal onset clusters sometimes develops into /th/ in thosedialects, so that in the word for 'fever,' the evolution /fjer/ > /fther/ >/ther/ is more likely. (Only the /f/ is phonemic in general Occitan.)"

References:(a) p.220 of M. J. Kümmel, Konsonantenwandel: Bausteine zu einerTypologie des Lautwandels und ihre Konsequenzen für dievergleichende Rekonstruktion. Wiesbaden: Reichert 2007.

(b) Point 9 of the Atlas Linguistique du Limousin et de l'Auvergne (LaChapelle-Marcousse (Puy-de-Dôme))----------(IV) /f/ ~ /theta/ dialect variation

Kelly Lynne Maynard reported that some sub-dialects of Albanian showconditioned /f/ ~ /th/ variation. The varieties in question are two NorthGheg dialects (Borgo Erizzo/Arbanasi and Peshter).

In Borgo Erizzo/Arbanasi, f > θ / _ t, e.g. [prift] > [priθt] 'priest'. InPeshter, Standard Albanian /ferra/ 'thorny bush' is realized as /therra/.This change is not due to a general ban on /fe/ sequences in Peshter:cf. Peshter /i fell/ vs. Standard Albanian /thell/ 'deep.'

All Albanian dialects have phonemic /f v θ ð/, so these changes are(potentially) neutralizing. Again, Kelly notes that the opposite shift (/th/> /f/) is quite common in Albanian dialects, as the above forms for'deep' illustrate.

Finally, Rémy Viredaz suggested that Tsakonian Greek underwent aconditioned /f/ > /th/ change. Rémy writes that "a change of /f/ to /theta/(sometimes /khi/) [occurred] in Tsakonian, however only before syllabicand asyllabic /i/ (from Ancient Greek /iota/ as well as /eta/, though notfrom /ypsilon/, which is known to have merged with /i/ only later, andnot before /e/); this change does not affect recent loans frommainstream Modern Greek."

Rémy points out that this change is likely linked to palatalization, as"Tsakonian also changes the sequences /ti/ and /pi/ to /ki/ (so written,but obviously with palatalized /k/; /k/ is palatalized before /e i/ inmodern Standard Greek)."

He also observes that this potential /f/ > /theta/ change is not the resultof a pull-chain or 'slot-filling' shift: though "Ancient Greek /theta/ shiftedto /s/ already in Antiquity (Laconian), Tsakonian neverthelesspossesses /theta/, e.g. /tha/ (future tense) or /tha’ssa/ (mainstreamGreek /tha’lassa/, ‘sea’)," independently of the /f/ > /th/ change.

References:(a) Weigand, Gustav. (1911) Der Gegische dialect von Borgo Erizzobei Zara in Dalmatien.

(b) Tagliavini, Carlo. (1937) L’albanese di dalmazia contributi allaconoscenza del dialetto Ghego di Borgo Erizzo presso Zara.

(c) p.104 of Latif Mulaku and Medhi Bardhi (1972) "Mbi të folmenShqipe të Peshterit," in Gjurmime Albanologjike.

(d) G.P. Anagnostopulos (1926). Tsakonische Grammatik, Berlin–Athen.

Once again, many thanks to those of you who replied. My deepestapologies if I have misrepresented any of your responses. Linguistic Field(s): Historical Linguistics                             Phonetics

Page Updated: 25-Jun-2010