[A phonemic analysis of non-weak-stressed syllables in the Peiping dialect is shown to yield three syllabic nuclei, three medials, and fifteen additional consonants. The problem of the ‘palatal series’ is investigated; the retroflex initials are shown to be clusters having a semivowel as the second element; it is concluded that there is only one pair of affricates in the dialect.]

[Koasati has certain well-defined differences between the speech of men and of women. Other Muskogean languages, such as Muskogee and Hitchiti, also had such differences formerly, and here and there throughout the world one meets with similar instances. Sometimes the differences affect vocabulary items, sometimes the pronunciation of particular words.]

To one who believes firmly that in the teaching of ancient as well as of modern languages an accurate pronunciation is the best possible start toward language mastery and the appreciation of literature, the pronunciation of Greek and Latin is a subject of unfailing interest. The evidence for the ancient pronunciation has been studied by a considerable number of scholars, and has recently been set forth excellently by Edgar H. Sturtevant in the second edition of his Pronunciation of Greek and Latin (Linguistic Society of America, 1940). The essential features of the phonology are, of course, settled beyond doubt, yet as to some lesser points of the pronunciation it is still possible to raise certain questions, even if one cannot always answer them. The theme of this discussion is the bearing of evidence from the modern languages on the pronunciation of gn in Latin and of γμ and γv in Greek.

[Despite the paucity of material, and notwithstanding the conflicting evidence, it would appear that Gaulish possessed a system of mutation like that of the other Celtic languages; and that this system resembled the Brythonic rather than the Goidelic. Certain analogous phenomena occur in French (and Engadine); but whether these are Gaulish survivals is a question, rather, for Romance scholars.]

Since Muratori, there have been several attempts to explain etymologically Ital. pozzánghera ‘mud-puddle’. The following ones stand out: 1. Caix identifies pozzánghera with Emilian pociacra and traces both back to Lat. *puteacula. 2. Meyer-Lübke considers it a derivative from puteus, but leaves it with the remark: ‘mit unerklärtem Ausgang’. 3. Gamillscheg sees the etymon in Langob. putz-angar ‘versumpfter Platz’, supporting his theory by geographical reasons. From the supposedly Langobardic origin of the compound pozzánghera Gamillscheg infers Langobardic origin also for the simple pozza ‘mud-puddle’. 4. Hubschmied derives both pozzánghera and pozzángola from puteus with the suffix -ang (which occasionally adds a pejorative nuance to the basic word), reinforced by a diminutive suffix. 5. Hall reconstructs on the basis of Emil. pociacra Ital. *pozzachera, according to him a derivative of pozza, enlarged by the pejorative -́era: pozzánghera, then, is a crossing of this *pozzachera with fango ‘mud’. The weakness of these hypotheses consists in the fact that they do not consider sufficiently either the older forms or the geographical distribution of the term under discussion.

By reference to Zipf's ‘standard curve of distribution’, we can conveniently describe the frequency distribution of words in any language. When the word-frequency curve for a given language is found to deviate markedly from the standard curve, then a consideration of such deviation may, as Zipf suggested, ‘shed welcome light on significant factors in the structure of the language.‘ By analysis of two vocabulary studies of French, I have found a French word-frequency curve which does in fact show definite peculiarities as compared with the standard. This paper is primarily to present that French curve for the consideration of linguists, and secondarily to suggest some approaches to ultimate interpretation of its significance.

In the oldest OE texts—the Epinal glossary, the Corpus glossary, and the oldest documents—b is written medially where other Germanic languages have /b/, and f is written where other Germanic languages have /f/; e.g. obær, halbæ, libr, hræbnes, ebor (Ep.), giaban (doc.), uulfes, cefr, hofr. The same distinction is found in final position; e.g. hualb, stæb, reab, uuf (Ep.), gib (doc.). In later OE texts, f is universally used both medially and finally; e.g. lifr, ofer, hræfn, efor, stæf, reaf. Some isolated spellings in Epinal indicated the coming change in spelling practice: unofercumenræ, sifunsterri, nabfogar. Corpus has even more cases, e.g. halfe, forscrifen, as well as reverse spellings like ceber, where b appears for f.

[A review of the theory that the Germanic weak preterite is derived from the 2d person singular middle of the Indo-European root-aorist.]

In spoken Finnish the imperative mode occurs in the 2d person singular, the 1st person plural, and the 2d person plural. In each form it is marked by one or more suffixes, and in spoken Finnish exhibits certain interesting features presenting a different picture from the one found in Finnish grammars, which do not generally deal with the colloquial forms.

According to the NED, the English noun runt is of obscure etymology. The editors rightly reject the connection with Middle Dutch runt ‘ox’ upheld (among others) by Weekley, but they offer nothing in its stead. H. Falk, in his Altnordische Waffenkunde 52, connects runt with Icelandic hrotti, a word which he defines as langer Mensch (longurio)'. In S. Blöndal's Islandsk-Dansk Ordbog the word hrotti is defined in Icelandic: ‘stór, klunnalega vaxinn maÐur (big, heavy, clumsy man)'; and in Danish: 'höj Mand af plump Vækst (tall, fat man)'. The derogatory sense 'coarse, rude fellow’ is recorded in Cleasby-Vigfússon's Icelandic-English Dict.; similarly in Blöndal and in Fritzner. In his Eigennamen im Beowulf 75, E. Björkman adds Danish runte, a form in which the sound-change nt > tt did not take place. In the Dansk Ordbog published under the auspices of Videnskabernes Selskab, runte is defined as 'en lang Stage, tyk ved den nederste Ende, til at springe over en Grøft med (a long pole, thick at the lower end, for use in jumping over a ditch)'. An Icelandic form without assimilation also occurs, though neither Falk nor Björkman mentions it. This form is hrunki, which Fritzner glosses ‘stor Person (big person)', while in Cleasby-Vigfússon only the derogatory senses 'clown, brute’ are recorded. According to Falk, runt and hrotti may go back to a stem *hrungt- (base hrung plus t-suffix); the base is cognate with OE hrung ‘rung, staff, rod, beam, pole,’ the modern rung (of a ladder or chair). I analyze hrunki as made up of this same base hrung plus k-suffix. The two words hrotti and hrunki obviously agree in meaning, whether used with or without derogation. They also agree in that they may be used as proper names. Hrunki is on record as a nickname, and Hrotti is on record as a sword-name: it occurs in the Elder Edda and the Völsungasaga as the name of Fáfnir's sword. Since the Scandinavians had a way of making a man's nickname do duty also as the name of his sword, the existence of the sword-name Hrotti indicates that Fáfnir had a nickname Hrotti, even though this nickname is not recorded as such. A nickname Hrotti would be eminently suitable for Fáfnir, who was by far the biggest and grimmest of the sons of HreiÐmarr and, when he exercised his powers as a shape-shifter, took the form of a huge dragon. To the sword-name Hrotti answers the sword-name Hrunting recorded in the OE poem Beowulf. This name shows the -ing-suffix regular in English (as distinguished from Scandinavian) sword-names. The sword in question presumably got its name from an owner nicknamed Hrunta. In other words, hrunta is the OE cognate of Icelandic hrotti, and the modern English runt can be traced back to OE times. Yet another Scandinavian proper name belongs here: the mythological name Hrungnir is obviously made up of a base hrung plus n-suffix, and since this name is that of a giant it goes well with the meaning ‘big person, tall man’ recorded for hrunki and hrotti. No Scandinavian form without a suffix occurs. Forms in other Germanic dialects without suffix, besides the OE hrung mentioned above, are Gothic hrugga ‘rod, staff’ and German runge.

The state of our etymological knowledge of this word is excellently summarized by REW §9439:

*VOLUS ‘fliegend’.

Afrz. veule ‘eitel’, ‘leer’, nfrz. veule ‘schlaff’, ‘weichlich’.—Zussg.: afrz. vanvole ‘nichtige Sache’, volgrain ‘Spreu’, volgrener ‘fliegen machen’.—Thomas, Mél. 214. (Zweifelhaft; VOLA Diez 698 ist begrifflich, dän. hvegl, schwed. vägel ‘unstet’ Vising, NTF. n.v. 7, 33 formell schwierig.)