BackgroundWith the rapid development of new psychoactive substances (NPS) and changes in the use of more traditional drugs, it is increasingly difficult for researchers and public health practitioners to keep up with emerging drugs and drug terms. Substance use surveys and diagnostic tools need to be able to ask about substances using the terms that drug users themselves are likely to be using. Analyses of social media may offer new ways for researchers to uncover and track changes in drug terms in near real time. This study describes the initial results from an innovative collaboration between substance use epidemiologists and linguistic scientists employing techniques from the field of natural language processing to examine drug-related terms in a sample of tweets from the United States.ObjectiveThe objective of this study was to assess the feasibility of using distributed word-vector embeddings trained on social media data to uncover previously unknown (to researchers) drug terms.MethodsIn this pilot study, we trained a continuous bag of words (CBOW) model of distributed word-vector embeddings on a Twitter dataset collected during July 2016 (roughly 884.2 million tokens). We queried the trained word embeddings for terms with high cosine similarity (a proxy for semantic relatedness) to well-known slang terms for marijuana to produce a list of candidate terms likely to function as slang terms for this substance. This candidate list was then compared with an expert-generated list of marijuana terms to assess the accuracy and efficacy of using word-vector embeddings to search for novel drug terminology.ResultsThe method described here produced a list of 200 candidate terms for the target substance (marijuana). Of these 200 candidates, 115 were determined to in fact relate to marijuana (65 terms for the substance itself, 50 terms related to paraphernalia). This included 30 terms which were used to refer to the target substance in the corpus yet did not appear on the expert-generated list and were therefore considered to be successful cases of uncovering novel drug terminology. Several of these novel terms appear to have been introduced as recently as 1 or 2 months before the corpus time slice used to train the word embeddings.ConclusionsThough the precision of the method described here is low enough as to still necessitate human review of any candidate term lists generated in such a manner, the fact that this process was able to detect 30 novel terms for the target substance based only on one month’s worth of Twitter data is highly promising. We see this pilot study as an important proof of concept and a first step toward producing a fully automated drug term discovery system capable of tracking emerging NPS terms in real time.

Hindko vocative case endings -97 4.3.3.3.2 Panjabi vocative case endings------98 4.3.3.3.3 Saraiki vocative case endings------99 4.3.3.4 Vocative particles------102 4.3.3.4.1 Hindko vocative particles------102 4.3.3.4.2 Panjabi vocative particles-----102 xiv -Contents 4.33.4.3 4.33.5 4.33.6 4.4 4.4.1 4.4.1.1 4.4.1.2 4.4.1.2.1 4.4.1.2.2 4.4.1.23 4.4.2 4.4.2.1 4.4.2.1.1 4.4.

This chapter makes two claims about Javanese, one concerning its internal dialect variation, and one concerning its place in mainland Southeast Asian (MSEA) typology. First, Javanese exhibits extreme dialect variation, with many features of these variants not appearing in descriptions of Javanese, which mostly concern the Central variety. Second, the existence of these features changes the position of Javanese in the continuum of isolating-to-synthetic languages. Relevant features from six dialects of Javanese show that the Central variety – that of Yogyakarta and Solo – inadequately characterises Javanese as a whole; rather, the geographically and socially ‘peripheral’ dialects more strongly tend toward isolating morphology. Consequently, Javanese is less of an outlier in the MSEA Sprachbund than is generally acknowledged. Historical evidence shows that the Central variety is innovative with respect to Javanese overall.

No abstract

Imposters, seemingly third person nouns with speech act participant reference, have been varyingly analyzed as being licensed through an elaborated DP syntax (Collins and Postal. 2008. Imposters. Manuscript. http://ling.auf.net/lingbuzz/000640 (accessed 12 May 2017), Collins and Postal. 2012. Imposters: A study of pronominal agreement. Cambridge: MIT Press) or through lexical specification (Kaufman 2014. The syntax of Indonesian imposters. In Chris Collins (ed.), Cross-linguistic studies of imposters and pronominal agreement, 89–120. Oxford: Oxford University Press). Looking at Korean and Indonesian, two languages that make frequent use of imposters, we show that both can be accounted for without appeal to an elaborated DP syntax and that, in fact, such a structure makes the wrong predictions. Rather, other heads in the clause, in conjunction with differences in lexical specification, can account for both languages. In Indonesian, which freely allows imposters to bind anaphors with person features of the referent, the imposter is lexically specified for those features. In Korean, where such binding is restricted, imposters are underspecified for person and so anaphors only occur when there is another person feature-carrying head to supply the necessary features (Zanuttini et al. 2012. A syntactic analysis of interpretive restrictions on imperative, promissive, and exhortative subjects. Natural Language & Linguistic Theory 30(4). 1231–1274). Previously left unexplained was why Korean imposters were unable to bind any person-marked anaphors, including third person, under an assumption that person-underspecified DPs get valued with a default third person feature. We argue this is a result of the difference in types of third person, those specified for third person and those that are not (Sigurðsson 2010. On EPP effects. Studia Linguistica 64(2). 159–189).