[The techniques of this study involve separating and recombining segments of utterances recorded on magnetic tape, and presenting the new stimuli to phonetically naive subjects for identification. The data support the results of previous experiments with synthetic speech conducted by the Haskins Laboratories: the place of articulation of m, n, and ŋ is conveyed principally by the transitions of the adjoining vowel formants. The nasal resonances, produced during the occlusion, serve principally as class markers, differentiating the nasals from their cognates in other classes, but they also appear to contribute a small amount to the place identification of these consonants.]

[In support of the theory that the Gothic digraph ai denoted relatively low front monophthongal values when it occurred in inflectional endings, the orthographic and morphological evidence is re-examined in detail. The paper shows that this evidence does not justify narrower quantitative and qualitative interpretations, and considers the possibility of reconciling the conclusions reached here with a plausible system of phonemic segmentation.]

[Spanish intonational structure is described in terms of three terminal junctures, three pitch phonemes, and three stress phonemes, with other intonation-like phenomena described as of a different order of structuring, the vocalizations. Internal open juncture is described as a segmental phoneme marked by a variety of distributional evidence in the allophones of the vocalic and consonantal segments.]

The existence of an ‘emphatic’ or velarized in Classical Arabic and the modern dialects has often been noted and the sound has been carefully described. Descriptions which have been made from a structural point of view have generally regarded this emphatic as an allophone of the usual l, not an independent phoneme, both in the Classical language and in the dialects. The purpose of this note is to make clear why the emphatic must be regarded as an independent phoneme in Classical Arabic and in most if not all the modern dialects.

1. Morris Swadesh, in a series of articles recently published, has described a method for dating the time of first separation of genetically related languages. The method, which he calls lexicostatistics or glottochronology, is said to provide ‘a time dimension for ethnic contacts shown by linguistic relationships’ (1955.121). The precision of this ‘time dimension’ is not easy to determine, but Swadesh claims (1955.123–4) that

[Samples of printed English sentences of three lengths were drawn randomly from a representative selection of popular magazines. The words of each sentence were classified according to Fries's system and a count was made of the various word classes at each sentence position. The tabulations were plotted as frequency distributions and treated statistically. Principally, the data obtained from all three sentence lengths indicate (1) that the greatest variations in word-class frequency tend to occur in sentence extremes and the immediately adjoining positions, and (2) that different word classes have characteristic patterns of variation.]

The relationship of Natchez to the Muskogean languages has been generally accepted for over thirty years. When Swanton first posited the relationship in 1924, he gave examples of lexical similarities between various individual Muskogean languages and Natchez; as a result of his postulation and his method of presenting the evidence, Natchez has often been referred to as a Muskogean language. Such nomenclature is, however, both misleading and inaccurate, because Natchez must be compared with Proto-Muskogean and not directly with the separate Muskogean languages. To indicate clearly the nature of the relationship that exists, it is more appropriate to speak of a Natchez-Muskogean stock composed of Natchez and the Muskogean family. Since no series of sound correspondences between reconstructed Proto-Muskogean (PM) and Natchez has ever been published, the purpose of the present paper is to make a preliminary statement concerning such correspondences. A brief description of the sounds of Natchez and of the Muskogean languages is provided first as background. This is followed by (1) a fist of possible cognates between Natchez and PM, (2) a list of the sound correspondences exhibited in these cognates, and (3) a discussion of certain other features which point to a relationship between the two languages.

In Mezquital Otomi certain phonological components (rather than phonemes as such) appear to differentiate between the aspect markers of the verbs. For example, the difference between the incompletive form ‘is or was scraping’ and the completive ‘did or will scrape (at a point of time)‘ is the component of voicing. Such components appear to have morphemic status in relation to the distribution of a large proportion of verb roots. The phonemic syllable, consisting most commonly of a consonant, a vowel, and a tone, is the matrix in which the interchange of components takes place. The syllable margin, specifically the initial consonant, is the focus of operation.

A. L. Kroeber has contributed to Hokan studies to an extent not usually recognized by nonspecialists. Together with Dixon, he first discovered the genetic relationships of the group, then consisting of Karok, Chimariko, Shastan (including Palaihnihan), Pomo, Yana, Esselen, and Yuman. He later added Seri, Tequistlatec (cf. Kroeber), and (with Dixon 1919) Washo. Kroeber has also contributed important fieldwork: Sapir's early comparative work leans exclusively on Kroeber's data for Chontal, Chumash, Salinan, and Esselen, and heavily on Kroeber's data for Yuman, Pomo, Karok, Shasta, Achumawi, and Atsugewi. Kroeber and other Hokanists have long noted the paucity of data from Achumawi and Atsugewi; the present paper is a result of field work undertaken at his suggestion.

For some time there has existed among some philosophers of language an assumption that linguistic development through time shows a universal trend from synthesis to analysis, from multimorphemic to monomorphemic words. Jespersen's theory of progress in language embodies such an assumption. Such hypotheses are based on preconceptions as to inferior and superior types of linguistic structures, with holophrasis at the inferior ‘primitive’ end of the continuum, and with languages which split such holophrases into monomorphemic words at the superior, ‘civilized’ end. Implicit likewise is the thesis that the superior type appeared concomitantly with development of high cultures. In contradistinction to this assumed trend of linguistic development, some evidence has been adduced for the opposite process, that of building up more complex words from earlier phrases, of a shift from analysis to synthesis. Swadesh made such a claim for Nootka in comparison with other related languages, and the isolated instances from Romance languages are well known.