Very few segments of the world's languages have been shown to have a systematic effect on the fourth formant (F4). We investigate a large drop in F4 which sometimes occurs in conjunction with the flap in American English. The goal of the present work is to document this phenomenon, and to determine what phonological environments coincide with this large drop in F4. We measure data from six speakers producing words with medial flaps in various environments, such as party, turtle, bottle, credit, harder. We find that the combination of flap with a rhotic and to a lesser extend a syllabic // leads to a larger drop in F4 than other flap combinations like a following /i/. Together with previous perceptual data, the findings support the conclusion that this feature of F4 results from transitions among articulations.

Phonetic notation of segmental sounds is of two main types, phonemic transcription and phonetic transcription proper which, to avoid ambiguity, is generally now referred to by British phoneticians, following the suggestion of Abercrombie (1953), as ‘allophonic transcription’. The two most important discussions of types of notation for English are Jones (1956: Appendix A) and Abercrombie (1964); for further comments see Windsor Lewis (1969: Chap. III) and Gimson (1970: § 5.6). Of the many different phonetic notations used for English more or less all of them in the last half century or so have employed the symbols of the Association's alphabet, in Britain usually faithfully observing the principles laid down by the Association for their use. In America the principles have not been so generally accepted. They are set forth in Jones, 1949 (PIPA (1949)).

The Baima language (/pêkê/, Chinese 白马语 báimăyŭ, ISO-639 code bqh) is a little-studied Tibeto-Burman (Bodic or Himalayish) language spoken in the southwestern part of the People’s Republic of China. Approximately 10,000 people, who traditionally reside in three counties in Sichuan Province (Pingwu 平武, Songpan 松潘 (in Written Tibetan, hereafter WT, zung chu), and Jiuzhaigou 九寨沟 (WT gzi rtsa sde dgu), and in one county in Gansu Province (Wenxian 文县), speak the Baima language (see Sichuan Sheng Minzu Yanjiusuo 1980, Zeng & Xiao 1987), see Figure 1.1 The largest concentrations of Baima speakers are in Baima Township (白马乡, Baima /tôpû/) of Pingwu County, and in Tielou Township (铁楼乡) of Wenxian County.

A uniform dialect of Canadian English (CE) encompasses a wider territory than that of any other regional variety (Priestley (1951); Woods (1979); cf. Scargill and Warkentyne (1972) for suggested subdivisions; cf. also Bernard (1969) versus Horvath (1985) for the scope of Australian English, another widespread variety). This is a result of converging influences in Canada of varieties of British English and of Northern and Midland American (von Baeyer 1977; Woods 1979). The components of CE are a distinctive body of lexical items marked foremost by compounding, with many borrowings from French and the native Indian languages (Avis 1973; Harris 1975; Gregg 1979), certain minor syntactic features along with the stereotypical use of ‘eh’ (Avis 1978; Bailey 1982; Chambers 1986), and a ‘General’ Canadian accent, recognized as urban and educated, spreading westward from Ontario to the Pacific, and affecting even eastern Maritime speech (Gregg 1984a; Avis 1986; cf. Kinloch 1983). Within this broad framework (Avis 1973, 1986; Gregg 1984a), however, certain social and regional distinctions appear when phonological variability is considered within the Labovian model of sociological co-variation (e.g. Labov 1966, 1972; Trudgill 1974; Milroy 1987). A comparison of phonological items from two recent and concurrent sociodialectal surveys, one in eastern Canada for Ottawa (Woods 1979) and the other on the Pacific Coast for Vancouver (Gregg 1984b), reveals certain points of phonetic divergence socially and regionally, together with differential rates of sound change (de Wolf 1988).

This article originated as a paper given to the Colloquium of British Academic Phoneticians in September 1978. That was based on an auditory-cum-proprioceptive investigation of some 200 ‘interesting voices’, taped from radio, TV and from some early jazz-records made available by Mr. P. Gras of Dutch radio.

This paper examines the phonetic realization of rising pre-nuclear pitch accents in Persian. In a first experiment, the alignment of f0 valleys and peaks in pre-nuclear pitch accents was analyzed in controlled speech materials as a function of the syllable structure (open vs. closed) and vowel type (short vs. long) of the accented syllable. The results revealed that in words with antepenultimate stress, both the L and the H tones are anchored to specific segmental landmarks irrespective of syllable structure or vowel type. In particular, the L is consistently aligned with the onset of the accented syllable, and the H is placed with similar consistency in the vicinity of the first post-accentual vowel. In a second experiment, the variability in the timing and scaling of L valleys and H peaks was examined as a function of the proximity of the word boundary and of the following accent. The results revealed that while the alignment of the L was unaffected by changes in stress conditions, H peaks were significantly retracted as the syllable approached the end of the word. However, the proximity of the following accent did not produce a significant effect on H alignment. In addition, no significant differences were found on L and H scaling in different stress or tonal crowding conditions. Overall, the results contribute to a growing body of evidence that in the absence of upcoming prosodic pressure, the alignment of pitch targets is specified relative to segmental positions. A comparison between these findings and empirical findings from other languages reveals fine phonetic differences of segmental anchoring that are less likely to be interpreted in terms of distinct association-based phonological representations, and suggests that some aspects of segmental anchoring need to be explained in terms of continuous language-specific alignment rules.

Diphthongization has a long history in Quebec French. In this variety, lengthened vowels have the potential to be diphthongized. However, one vowel stands out, as it is always long and seems particularly inclined to be diphthongal: the fête vowel. While it has been much studied over the years, some uncertainties remain in the literature, especially whether length or quality primarily distinguishes it and which phonetic transcription appropriately reflects the way Quebec French speakers pronounce it today. This study addresses these issues by drawing a parallel between acoustic properties of the fête vowel and those of other vowels produced in similar consonantal contexts by 52 native speakers of Quebec French. After analyzing 8866 tokens, we present a descriptive account of their relative duration, location in F1/F2 planes and spectral changes. The results show that fête is, in fact, acoustically distinct, but more in terms of the spectral changes that occur than its duration. Its first two formants extensively shift between 25% and 75% of its duration, with a substantial F2 increase and F1 decrease. In F1/F2 planes, fête has onset values similar to /a/ and offset values similar to /ɛ/. Therefore, we argue that the two most commonly used phonological transcriptions for this singular vowel, /ɜ/ and /ɛː/, present some problems and do not capture its characterizing features. Drawing on our results and the IPA vowel chart, we suggest a more accurate symbol.

This study focuses on a corpus assembled from commercial recordings of 1929 and 1932 featuring the two British linguists J. R. Firth (1890–1960) and Daniel Jones (1881–1967). The aim is to analyse and quantify differences in the Received Pronunciation (RP) used by the two men, in relation to sociolinguistic and stylistic variation within RP of the period. A systematic acoustic analysis reveals that whereas the vowel systems of the two speakers are closely similar in most respects, there are significant differences in the realizations of the trap, price and dress qualities. We show that each of these has a well-documented history of variation in RP, and find that in each case Firth is employing the higher-class or more prestigious variant, which is a reversal of expectations based on what is known about the social and regional origins of the two. We consider the possible roles of social background, avoidance of regional features and hypercorrection. The outcomes of our work are (i) an illustration of RP used in the interwar period for the purposes of teaching English as a foreign language, (ii) empirical evidence for internal variation within the accent, (iii) additional insights into the stylistic and social correlates of this variation, and (iv) a demonstration that satisfactory formant analyses can be conducted with recordings from as early as the 1920s. Overall, it is hoped that this case-study will both throw light on the ‘standard’ accent of the era and inspire further sociophonetic investigations with legacy recordings.

Cochabamba Quechua disallows pairs of ejectives within roots (*[k’it’ɑ]), but this structure may arise across word boundaries, e.g. [misk’it’ɑntɑ] ‘good bread’. This paper presents an acoustic study of these phonotactically legal, trans-vocalic ejective pairs that occur at word boundaries. It is found that Cochabamba Quechua speakers de-ejectivize one of the two ejectives in such phrases a significant portion of the time, and that, in correct productions with two ejectives, the period between the two ejectives is lengthened by increasing the duration of the vowel and the closure duration of the second ejective.

The IPA alphabet is, first and foremost, a device for use in the making of impressionistic records of spoken language events. There is no alternative device in widespread use. It is neither full nor systematic as a device for this purpose, but can be used remarkably effectively if supplemented with ad hoc extensions and adaptations. Such adaptations need to be explicit if used in published material, but they need not be standard. They do need, though, to be sensible. Although the IPA letter-shapes are not systematic, the articulatory classification that underlies them is. Within this classification it would be perverse to relate a ‘linguo-labial’ to, say, the category ‘velar’, and so a representation based on the letter-shape ‘k’ would be undesirable. Given these limits, though, I can see no theoretical reason why ‘linguo-labials’ should not be written in a number of ways. [P‥] etc. is one way. It is not a particularily good one, in my view, as [‥] is a diacritic that distinguishes a PASSIVE articulator, and this passive articulator is nowhere involved in ‘linguo-labials’. But I would happily accept [P‥]. It would be desirable for a degree of uniformity to apply in PUBLISHED material, but this is a matter more for editors and for publishers than for front-line linguists. At the PHONETIC level, I can't see that it matters whether we use [P‥] or, say, [‡] in published material: and other suggestions will no doubt be forthcoming, PRACTICAL considerations are paramount here, some solutions being easier to print, type and read than others.

Labellers of connected speech corpora encounter a number of cases which are difficult to handle with a linear sequence of segmental labels. Based on material from a corpus of German unscripted speech, this paper presents non-sequential aspects of laryngeal activity for /x/ and /h/, and of supralaryngeal articulation in vocoid exponents of postvocalic /r/. Comparative data for /r/ from a corpus of German read speech indicate that differences across speaking styles may be independent of durational variation. It is discussed how models of the phonology-phonetics interface and Prosodic Analysis can account for the findings.

Seenku (ISO 639-3: sos) is a Western Mande language of the Samogo group, whose other members include languages like Dzùùngoo (Solomiac 2014), Jowulu (Djilla, Eenkhoorn & Eenkhoorn-Pilon 2004), and Duungooma (Hochstetler 1996), spoken on either side of the Mali-Burkina Faso border. The endonymic language name Seenku [sɛ̃́ː-kû] (also spelled on Ethnologue as Seeku) literally means ‘thing of the Sɛ̃ː ethnicity’, but it is widely known to outsiders as Sembla (variant spelling Sambla), which doubles as an exonym for the ethnicity. Seenku has two primary dialects, Northern and Southern, spoken in villages approximately 40 km west of Bobo-Dioulasso in Burkina Faso (see map in Figure 1). This study focuses on the more populous southern dialect, particularly the variety spoken in and around the large village center of Bouendé (local name [ɡ͡béné-ɡũ̏]), with a population of approximately 12,000 speakers; the Northern dialect, spoken around the village center of Karangasso (local name [təmî]), has a population of approximately 5000 speakers and was the subject of a sketch grammar (Prost 1971). The southern dialect had until recently received little scholarly attention, with the exception of a Master's thesis on the morphophonology at the Université de Ouagadougou (Congo 2013), but is now the subject of the NSF Documenting Endangered Languages grant supporting this research (BCS-1664335). Other published work includes McPherson (2017a, b, c, d).

The following suggestions emerge from my paper read at the Eighth International Congress of Phonetic Sciences and recapitulate proposals made at the meeting of the International Phonetic Association held in connection with that Congress.

Taba (also known as ‘East Makian’ or ‘Makian Dalam’) is an Austronesian language spoken in northern Maluku province, Indonesia. It is spoken on Makian, Moti and Kayoa islands and also in a few villages on other nearby islands. The speech recorded here is from the village of Ngofakiaha on Makian island. Slight dialectal variations are to be found in most villages where the language is spoken.

Brazilian Veneto is a Romance language spoken by approximately 500,000 people, most of whom live in the southern Brazilian states of Rio Grande do Sul, Santa Catarina and Paraná (Margotti 2004, Pereira 2017). It is also spoken in the southeastern Brazilian states of São Paulo and Espírito Santo (see e.g. Loriato 2019). The language is often referred to by its speakers as Talian (/taˈljaŋ/), meaning ‘Italian’, given the geographical origin of the speakers who developed it, even though it is not a variety of Standard Italian. In this Illustration, I will refer to the language under examination as Talian, following many speakers’ naming preference.

Traditional analyses of French recognize three nasal consonant phonemes /m n ɲ/ appearing in the labial, apical and palatal orders respectively. What is interesting about this situation is that while French contains labial and apical oral stops /p b t d / corresponding to /m/ and /n/, there are no palatal occlusives linked to /ɲ/. Conversely, in the velar order where we find /k/ and /g/, there is no velar nasal representative. Thus, in a widely used presentation, Malmberg notes that

‘Le français n'a pas de phonème /ɲ/. Un son [ɲ] peut exceptionnellement apparaître comme variante par suite d'une assimilation de /g/ en contact avec une nasal: une longue minute [yn1Ʒɲminyt]. Dans les emprunts anglais comme smoking, parking, camping, il semble qu'un nombre croissant de Français prononcent une nasale vélaire /ɲ/.’

(1969: 106).

One can also cite the words of Fouché (1959: XIX), who notes that,

‘Il faut aussi ajouter à cette liste [de consonnes françaises] le son [ɲ] (= n vélaire) de la terminaison anglaise -ing qui a pénétré par voie d'emprunt dans le système consonantique français; cf. meeting, parking, smoking, etc.’

Neither author, however, explores the question in any detail.

Sardinian is a Romance language spoken almost exclusively on the island of Sardinia, an autonomous region of Italy. Sardinian and Italian are not mutually intelligible; there is considerable structural distance between the two linguistic systems, at all linguistic levels (Loporcaro 2009: 162–171).

Dàgáárè /dàgááɹɪ̀/ (ISO 639-3) is a Mabia language (Bodomo 1997) of Niger-Congo family. It is spoken by about 1.5 million The map in Figure 1 shows the areas, in northwestern Ghana and southern Burkina Faso, where Dàgáárè is spoken predominantly. There are four broad dialects of Dàgáárè. These include Northern Dàgáárè [dàgàɹà], spoken in Lawra /lóɹáː/, Nandom /nàndɔ̀ː/ and Burkina Faso; Southern Dàgáárè /wáːlɪ́/, spoken around Kaleo /kàlèó/ and in Wa /wá/; Western Dàgáárè /bɪ̀ɹɪ̀fɔ̀/ spoken in Lassie /lààsɪ̀ɛ́/, Tuna /tʊ́ːnà/ and along the western side of the Black Volta river in Burkina Faso. The other dialect is Central Dàgáárè. These broad dialects are further divided into sub-dialects, as there are internal variations in these dialect groups (Bodomo 2000). This paper’s primary focus is on Central Dàgáárè, which comprises the varieties spoken in and around Jirapa /ʤɪ̀ɹɛ́bǎː/, Han /hɛ̌ŋ/, Ullo /úlò/, Daffiama /dàfɪ̀ɛ́mɛ́/, Nadowli /nàdòlí/, Charikpong /tʃɛ̀ɹɪ̀kpóŋ/, Sombo /sʊ̀mbɔ́/ and Duong /dùóŋ/.

The appearance of the Journal of the International Phonetic Association offers a natural opportunity for taking both a look backward and a look forward in the field of phonetics. Ever since the publication of The Phonetic Teacher in 1886 and through the long period of Le Maître Phonétique, the organ of the Association has continued to appear, even if during war years the interval between successive numbers may sometimes have been long. It now comes in a new guise, the major change being the setting of the body of the journal in orthography instead of in phonetic transcription, and this is a suitable moment at which to ask what this change, a far from superficial one, betokens: what kind of scholars were the men who formed the Association and guided its early years, was their work of lasting value, what fresh light has been shed on the field in the intervening period, where do we stand now and what guesses can we make about the sort of work our successors will be doing?