Linguistic Constraints on Statistical Word Segmentation: The Role of Consonants in Arabic and English

Abstract

Statistical learning is often taken to lie at the heart of many cognitive tasks, including the acquisition of language. One particular task in which probabilistic models have achieved considerable success is the segmentation of speech into words. However, these models have mostly been tested against English data, and as a result little is known about how a statistical learning mechanism copes with input regularities that arise from the structural properties of different languages. This study focuses on statistical word segmentation in Arabic, a Semitic language in which words are built around consonantal roots. We hypothesize that segmentation in such languages is facilitated by tracking consonant distributions independently from intervening vowels. Previous studies have shown that human learners can track consonant probabilities across intervening vowels in artificial languages, but it is unknown to what extent this ability would be beneficial in the segmentation of natural language. We assessed the performance of a Bayesian segmentation model on English and Arabic, comparing consonant-only representations with full representations. In addition, we examined to what extent structurally different proto-lexicons reflect adult language. The results suggest that for a child learning a Semitic language, separating consonants from vowels is beneficial for segmentation. These findings indicate that probabilistic models require appropriate linguistic representations in order to effectively meet the challenges of language acquisition.

Does morphological complexity affect word segmentation? Evidence from computational modeling

“…Computational modeling work has started to investigate word segmentation in various languages (Batchelder, 2002;Blanchard et al, 2010;Caines, Altmann-Richer & Buttery, 2019;Daland, 2009;Fleck, 2008;Fourtassi, Börschinger, Johnson & Dupoux, 2013;Kastner & Adriaans, 2017;Saksida et al, 2017). Providing a thorough overview of their findings is beyond the scope of the present study, but we would like to highlight that most previous work attempts to check how a given algorithm performs cross-linguistically to argue for the validity of the algorithm the authors of those studies proposed, rather than to understand whether language properties affect segmentation in a systematic way (e.g., Batchelder, 2002;Boruta, Peperkamp, Crabbé and Dupoux, 2011;M.…”

“…Johnson, 2008;Pearl and Phillips, 2018;Phillips and Pearl, 2014a). Exceptions include studies that try to explain away cross-linguistic differences on the basis of corpus characteristics (e.g., Caines et al, 2019;Fourtassi et al, 2013), and work assessing the effect of prosodic and syntactic structure such as head direction (saliently, Gervain and Erra, 2012;Saksida et al, 2017), or the effects of input representation (Kastner & Adriaans, 2017). However, these factors are orthogonal to the present study (i.e., they are not necessarily confounded with morphological complexity).…”

xMatcher: Matching Extensible Markup Language Schemas using Semantic-based Techniques

“…For matching results, we used the previously published results produced by twelve ontology matching systems (SANOM [31], AML [13], LogMap [32], XMap [33], KEPLER [34], ALIN [9], DOME [11], Holontology [10], FCAMapX [35], [36], LogMapLt [32], ALOD2Vec [12], and Lily [37]) that [38] to evaluate the returned matches based on nine combinations of evaluation variants with crisp reference alignments: ra1-M1, ra1-M2, ra1-M3, ra2-M1, ra2-M2, ra2-M3, rar2-M1, rar2-M2, and rar2-M3 (ra1 is the original reference alignment; ra2 is an extension of ra1; and rar2 is an updated version of ra2 that deals with violations of conservativity). ra1-M1, ra2-M1, and rar2-M1 are used to evaluate only alignments between classes; ra1-M2, ra2-M2, and rar2-M2 are used to evaluate only alignments between properties; and ra1-M3, ra2-M3, and rar2-M3 are used to evaluate both alignments between classes and properties.…”

Templatic morphology as an emergent property

“…In recent computational work, Kastner and Adriaans (2017) have proposed that if the learner divides the input into consonants and vowels, it should be able to make progress on basic acquisition tasks in Semitic. A number of computer simulations found that ignoring the vowels in the input leads the learner to perform better on the task of segmenting the input stream into separate phonological words in Arabic: presumably, if the input consisted only of root consonants (and the occasional affix or clitic), this situation would lend itself surprisingly well to insertion of word boundaries.…”