Variation and Change in French Morphosyntax: The Case of Collective Nouns

Collective nouns such as majorité or foule have long been of interest to linguists for their unusual semantic properties, and provide a valuable source of new data on the evolution of French grammar. This book tests the hypothesis that plural agreement with collective nouns is becoming more frequent in French. Through an analysis of data from a variety of sources, including sociolinguistic interviews, gap-fill tests and corpora, the complex linguistic and external factors which affect this type of agreement are examined, shedding new light on their interaction in this context. Broader questions concerning the methodological challenges of studying variation and change in morphosyntax, and the application of sociolinguistic generalisations to the French of France, are also addressed.

Reviews:

‘Cet ouvrage constitue un apport majeur dans le champ de la linguistique variationniste et diachronique, tant par les résultats mis au jour que par la qualité de sa démarche méthodologique.’ — Sophie Prévost, French Studies 69.4, October 2015, 578-79

‘While language variation and change have been the focal point for linguists on this side of the Atlantic, Tristram argues that studies on morphosyntactic variation in French studies are lacking due to a focus on phonology and dialectology as well as denial of variation and change in the French language. Tristram’s book is thus a welcome contribution.’ — Samira Hassa, French Review 89.3, 2016, 108

‘Anyone teaching variation in French will want to talk about the findings and reflections reported in this study. A remarkable amount of ground is covered in a small compass. This is a highly welcome addition to the Legenda list, and one must hope that further linguistics titles will be added to it before very long.’ — Nigel Armstrong, Journal of French Language Studies 26.2, 2016, 211-13

L'accord sujet-verbe en français contemporain: une étude de variation sociolinguistique

Cette communication a pour objectif de présenter et d’analyser les résultats d’une enquête portant sur l’accord sujet-verbe en français contemporain. Dans le domaine de l’accord sujet-verbe, bien que dans la plupart des cas le locuteur n’ait pas le choix de l’accord – c’est–à-dire qu’il n’y a qu’un accord possible – il existe néanmoins des contextes dans lesquels on peut trouver une variation entre l’accord singulier et le pluriel (cf. Corbett 2006 ; Grevisse 1993 ; Riegel et al 1994). Cette variation est souvent liée à une discordance entre le nombre syntaxique et le nombre sémantique. C’est le cas de certaines expressions de quantité, comme dans les exemples suivants : Singulier : « On a affaire à une minorité qui fait la loi à l’université » (Ouest France, 23-24 mai 2009, p.13) Pluriel : « Un petit millier de producteurs allemands, français et belges se sont déplacés, hier, à Bruxelles […] » (Ouest France, 26 mai 2009, p.3) Cette variation nous offre plusieurs pistes de recherche : dans une perspective linguistique, elle peut nous aider à mieux comprendre comment interagissent les différents facteurs linguistiques qui ont une influence sur l’accord, et dans une perspective sociolinguistique, elle représente un nouveau domaine à explorer pour l’étude sociolinguistique de la variation grammaticale en français, ce qui reste jusqu’à présent relativement peu étudiée. Nous traitons dans cette communication de la perspective sociolinguistique, c'est-à-dire les facteurs externes tels que l’âge, le sexe, et le niveau d’éducation du locuteur qui jouent un rôle dans l’accord sujet-verbe avec les expressions de quantité. Nous considérons en particulier la variation sexolectale : dans un premier temps, nous examinons les résultats de quelques études précédentes de la variation morphosyntaxique en français contemporain par rapport à l’influence du sexe du locuteur. Nous en concluons que les Principes élaborés par Labov (1990) pour décrire la variation sexolectale en anglais semblent être moins valables pour le cas du français de la France ; ou bien, qu’ils ne s’appliquent pas de façon simpliste. Dans un deuxième temps, nous présentons les résultats de nôtre étude, et nous voyons que pour nôtre projet aussi, les résultats pour la variation sexolectale ne s’expliquent pas facilement dans le cadre des Principes de Labov (1990).

A genome-wide linkage scan for genes controlling variation in urinary albumin excretion in type ll diabetes

The main hallmark of diabetic nephropathy is elevation in urinary albumin excretion. We performed a genome-wide linkage scan in 63 extended families with multiple members with type II diabetes. Urinary albumin excretion, measured as the albumin-to-creatinine ratio (ACR), was determined in 426 diabetic and 431 nondiabetic relatives who were genotyped for 383 markers. The data were analyzed using variance components linkage analysis. Heritability (h2) of ACR was significant in diabetic (h2=0.23, P=0.0007), and nondiabetic (h2=0.39, P=0.0001) relatives. There was no significant difference in genetic variance of ACR between diabetic and nondiabetic relatives (P=0.16), and the genetic correlation (rG=0.64) for ACR between these two groups was not different from 1 (P=0.12). These results suggested that similar genes contribute to variation in ACR in diabetic and nondiabetic relatives. This hypothesis was supported further by the linkage results.

Psychotic symptoms in Alzheimer's disease are not influenced by polymorphic variation at the dopamine receptor DRD3 gene

It has been suggested that genetic influences unmasked during neurodevelopment to produce schizophrenia may appear throughout neurodegeneration to produce AD plus psychosis. Risk of schizophrenia and psychosis in Alzheimer's disease (AD) has been linked to polymorphic variation at the dopamine receptor DRD3 gene implying similar causative mechanisms. We tested this association in a large cohort of Alzheimer's disease patients with a diagnosis of probable AD of 3 years or more duration from the relatively genetically homogenous Northern Irish population. We assessed relationships between genotypes/alleles of the DRD3 BalI polymorphism and the presence or absence of psychotic symptoms (delusions, hallucinations) in AD patients during the month prior to interview and at any stage during the dementia. No significant associations were found when delusions and hallucinations were cross-tabulated against S and G alleles and SS, SG and GG genotypes. Logistic regression failed to detect any influence of APOE, gender, family history or prior psychiatric history. In conclusion, we were unable to confirm previously reported associations between the DRD3 BalI polymorphism and psychotic symptoms in AD.

Sequence variation in the CHAT locus shows no association with late-onset Alzheimer's disease

There is substantial evidence for a susceptibility gene for late-onset Alzheimer's disease (AD) on chromosome 10. One of the characteristic features of AD is the degeneration and dysfunction of the cholinergic system. The genes encoding choline acetyltransferase (ChAT) and its vesicular transporter (VAChT), CHAT and SLC18A3 respectively, map to the linked region of chromosome 10 and are therefore both positional and obvious functional candidate genes for late-onset AD. We have screened both genes for sequence variants and investigated each for association with late-onset AD in up to 500 late-onset AD cases and 500 control DNAs collected in the UK. We detected a total of 17 sequence variants. Of these, 14 were in CHAT, comprising three non-synonymous variants (D7N in the S exon, A120T in exon 5 and L243F in exon 8), one synonymous change (H547H), nine single-nucleotide polymorphisms in intronic, untranslated or promoter regions, and a variable number of tandem repeats in intron 7. Three non-coding SNPs were detected in SLC18A3. None demonstrated any reproducible association with late-onset AD in our samples. Levels of linkage disequilibrium were generally low across the CHAT locus but two of the coding variants, D7N and A120T, proved to be in complete linkage disequilibrium.

Plasma Homocysteine Is Not Subject to Seasonal Variation.

Background: Studies investigating the relationship between plasma total homocysteine (tHcy) and vascular disease usually rely on a single measurement. Little information is available, however, on the seasonal variability of plasma tHcy. The aim of this study was to investigate the seasonal variation in fasting plasma tHcy and related B-vitamin intake and status in a group of people who did not consume fortified foods or take B-vitamin supplements. Methods: In this longitudinal study, a group of 22 healthy people were followed for 1 year. A fasting blood sample and dietary information were collected from each individual every 3 months, i.e., at the end of each season. Results: There was no significant seasonal variation in plasma tHcy or in B-vitamin intake or status with the exception of red cell folate (significantly lower in spring compared with autumn or winter) and serum folate (significantly lower in spring compared with the other seasons). Although the between-person variation in plasma tHcy was high (47%), the within-person variation was low (11%). This low variation, combined with the low methodologic imprecision of 3.8%, yielded a high reliability coefficient for plasma tHcy (0.97). Conclusions: Although there was a small seasonal variation in folate status, there was no corresponding seasonal variation in plasma tHcy. The high reliability coefficient for plasma tHcy suggests that a single measurement is reflective of an individual’s average plasma tHcy concentration, thus indicating its usefulness as a potential predictor of disease. This, however, needs to be confirmed in different subgroups of the population.

A reliable externally fixated murine femoral fracture model that accounts for variation in movement between animals.

Fifty-two CFLP mice had an open femoral diaphyseal osteotomy held in compression by a four-pin external fixator. The movement of 34 of the mice in their cages was quantified before and after operation, until sacrifice at 4, 8, 16 or 24 days. Thirty-three specimens underwent histomorphometric analysis and 19 specimens underwent torsional stiffness measurement. The expected combination of intramembranous and endochondral bone formation was observed, and the model was shown to be reliable in that variation in the histological parameters of healing was small between animals at the same time point, compared to the variation between time-points. There was surprisingly large individual variation in the amount of animal movement about the cage, which correlated with both histomorphometric and mechanical measures of healing. Animals that moved more had larger external calluses containing more cartilage and demonstrated lower torsional stiffness at the same time point. Assuming that movement of the whole animal predicts, at least to some extent, movement at the fracture site, this correlation is what would be expected in a model that involves similar processes to those in human fracture healing. Models such as this, employed to determine the effect of experimental interventions, will yield more information if the natural variation in animal motion is measured and included in the analysis.

Assocation Between Variation in the Actin-Binding Gene Caldesmon and Diabetic Nephropathy in Type 1 Diabetes

Dysfunction of the actin cytoskeleton is a key event in the pathogenesis of diabetic nephropathy. We previously reported that certain cytoskeletal genes are upregulated in mesangial cells exposed to a high extracellular glucose concentration. One such gene, caldesmon, lies on chromosome 7q35, a region linked to nephropathy in family studies, making it a candidate susceptibility gene for diabetic nephropathy. We screened all exons, untranslated regions, and a 5-kb region upstream of the gene for variation using denaturing high-performance liquid chromatography technology. An A>G single nucleotide polymorphism (SNP) at position -579 in the promoter region was associated with nephropathy in a case-control study using 393 type 1 diabetic patients from Northern Ireland (odds ratio [OR] 1.38, 95% CI 1.02–1.86, P = 0.03). A similar trend was found in an independent sample from a second center. When the sample groups were combined (n = 606), the association between the -579G allele and nephropathy remained significant (OR 1.35, 1.07–1.70, P = 0.01). The haplotype structure in the surrounding 7-kb region was determined. No single haplotype was more strongly associated with nephropathy than the -579A>G SNP. These results suggest a role for the caldesmon gene in susceptibility to diabetic nephropathy in type 1 diabetes.