The relationship between a history of substance use disorder and the early course of psychotic illness was examined in 96 subjects with schizophrenia and 106 subjects with affective psychosis followed in the Suffolk County Mental Health Project, a longitudinal study of first-admission psychosis. Subjects received a structured diagnostic interview and clinical ratings at baseline assessment and again 6 months later. The 6-month assessment included information about treatment received during the interval. A lifetime history of substance use disorder was associated with worse clinical functioning at 6 months for schizophrenia subjects, but not for those with affective psychosis. There were no significant associations of substance use disorder with type of treatment during the interval or with self-reported compliance with medication. Schizophrenia subjects were more likely than subjects with affective psychosis to report cannabis use during the interval and to meet criteria for cannabis use disorder.

This article describes the rationale, aims, and methodology of an epidemiological study of psychosis being conducted in Suffolk County, New York. A sample of first-admission patients is drawn from 10 inpatient and 25 outpatient facilities. Diagnostic psychosocial interviews are conducted shortly after admission to treatment, and at 6- and 24-month followup. Consensus diagnoses are made after each interview. Demographic and clinical background characteristics of the first 250 subjects enrolled over a 2-year period are presented here. The response rate was 76 percent. Based on the initial interview, 75 percent of subjects received a diagnosis involving psychosis. The three most common diagnoses were schizophrenia, bipolar disorder with psychotic features, and major depression with psychotic features. Among subjects with psychosis, 58 percent of males and 29 percent of females had a history of substance abuse/dependence. Gender differences were found on several background and clinical characteristics. Males were somewhat younger, less likely to have ever married, and had less education. Although the median length of hospitalization was the same for females and males (27 days), females were more likely to be hospitalized within 1 month of the occurrence of their first psychotic symptom (60% of females compared to 37% of males). Subjects with schizophrenia-related disorders were significantly more impaired on an assessment of negative symptoms than were affectively ill subjects, but clinical ratings of depression were not significantly different across diagnostic groups.

This paper describes the consonant inventory of the endangered southern African language Nǀuu. Our novel approach to segment classification accounts for all 73 Nǀuu consonants with just four phonetic dimensions (place, manner, phonation, airstream) and does away with the phonetically empty categoryclick accompaniment. We provide ultrasound data showing that the posterior constrictions in clicks are not produced at the ‘velar’ place of articulation, and that posterior place differs with anterior place. We therefore argue for a terminological shift fromvelarictolingualairstream mechanism. Our data also show that the posterior place of articulation is the same in Nǀuu's five lingual ([⊙ ǀ ǃ ǁ ǂ]) and linguo-pulmonic ([]) stops. We argue that the difference between these segment classes is best captured in terms of airstream, not place. Plain clicks use only the lingual airstream, while linguo-pulmonic segments are airstream contours, in which the transition to the pulmonic airstream occurs within the segment rather than at its boundary. Our evidence suggests that the contrast between ‘velar’ and ‘uvular’ clicks proposed for the related language ǃXóõ is likely also one of airstream and that a contrast solely in terms of posterior place would be articulatorily impossible.

Ju|'hoansi contains parallel glottalization, breathiness and epiglottalization on consonants and vowels, as well as uvularization on consonants. There are voice quality cues associated with epiglottalized vowels that parallel those found with breathy and glottalized vowels. There is also guttural coarticulation, whereby aspirated, glottalized, uvularized and epiglottalized consonants spread acoustic voice quality cues onto a following vowel. The laryngeal coarticulation exhibited by click consonants is shown to be largely analogous to coarticulation found with pulmonic consonants. Two levels of acoustic similarity involving gutturals are identified. Low HNR in the C-V transition unites all gutturals, and low spectral slope differentiates glottals and epiglottals from aspirates and uvularized sounds that are associated with high spectral slopes. The two natural classes identified by these cues are shown to be targeted in perceptuallymotivated sound patterns. In addition to the four-way phonation type contrast on vowels, there are also diphthongs in breathiness and epiglottalization, leading to a three-way timing contrast in roots. Sufficient temporal separation of voice quality cues is not achieved between roots containing similar guttural consonants and guttural vowels. Phonotactic patterns assure that similar consonants and vowels do not co-occur within a single root, allowing Voice Onset Time to assure sufficient discriminibility of roots containing guttural consonants and roots containing guttural vowels. This suggests that sound inventories can not be evaluated for discriminability in isolation of existing cooccurrence patterns. * I would like to thank Mary Beckman for her help in designing, implementing and analyzing the results of the experiments reported on here, which were originally included in my dissertation under her direction (Miller-Ockhuizen 2001). I would also like to thank Abby Cohn, Osamu Fujimura, Bonny Sands, Marisol del Teso-Craviotto, and members of my seminar on the acoustics of voice quality taught at Cornell in the spring of 2002 (Marc Brunelle, Rina Kreitman and Anastasia Riehl) for comments on a draft version of this paper.