Measuring sign complexity: Comparing a theory-driven and a data-driven approach

The study of articulatory complexity has proven to yield useful insights into the phonological mechanisms of spoken languages. In sign languages, this type of knowledge is scarcely documented. The current study compares a data-driven measure and a theory-driven measure of complexity for signs in French Sign Language (LSF). The former measure is based on error rates of handshape, location, orientation, movement and sign fluidity in a repetition task administered to non-signers; the latter measure is derived by applying a feature-geometry model of sign description on the same set of signs. A significant correlation is found between the two measures for the overall complexity. When looking at the impact of individual phonemic classes on complexity, a significant correlation is found for handshape and location but not for movement. We discuss how these results indicate that a fine-grained theoretical model of sign phonology/phonetics reflects the degree of complexity as resulting from the perceptual and articulatory properties of signs.

Exploring Networks of Lexical Variation in Russian Sign Language

When describing variation at the lexical level in sign languages, researchers often distinguish between phonological and lexical variants, using the following principle: if two signs differ in only one of the major phonological components (handshape, orientation, movement, location), then they are considered phonological variants, otherwise they are considered separate lexemes. We demonstrate that this principle leads to contradictions in some simple and more complex cases of variation. We argue that it is useful to visualize the relations between variants as graphs, and we describe possible networks of variants that can arise using this visualization tool. We further demonstrate that these scenarios in fact arise in the case of variation in color terms and kinship terms in Russian Sign Language (RSL), using a newly created database of lexical variation in RSL. We show that it is possible to develop a set of formal rules that can help distinguish phonological and lexical variation also in the problematic scenarios. However, we argue that it might be a mistake to dismiss the actual patterns of variant relations in order to arrive at the binary lexical vs. phonological variant opposition.

Vision-based Pakistani Sign Language Recognition Using Bag-of-Words and Support Vector Machines

In order to perform their daily activities, a person is required to communicating with others. This can be a major obstacle for the deaf population of the world, who communicate using sign languages (SL). Pakistani Sign Language (PSL) is used by more than 250,000 deaf Pakistanis. Developing a SL recognition system would greatly facilitate these people. This study aimed to collect data of static and dynamic PSL alphabets and to develop a vision-based system for their recognition using Bag-of-Words (BoW) and Support Vector Machine (SVM) techniques. A total of 5,120 images for 36 static PSL alphabet signs and 353 videos with 45,224 frames for 3 dynamic PSL alphabet signs were collected from 10 native signers of PSL. The developed system used the collected data as input, resized the data to various scales and converted the RGB images into grayscale. The resized grayscale images were segmented using Thresholding technique and features were extracted using Speeded Up Robust Feature (SURF). The obtained SURF descriptors were clustered using K-means clustering. A BoW was obtained by computing the Euclidean distance between the SURF descriptors and the clustered data. The codebooks were divided into training and testing using 5-fold cross validation. The highest overall classification accuracy for static PSL signs was 97.80% at 750×750 image dimensions and 500 Bags. For dynamic PSL signs a 96.53% accuracy was obtained at 480×270 video resolution and 200 Bags.

Analytical review of models and methods for automatic recognition of gestures and sign languages

Introduction: Currently, the recognition of gestures and sign languages is one of the most intensively developing areas in computer vision and applied linguistics. The results of current investigations are applied in a wide range of areas, from sign language translation to gesture-based interfaces. In that regard, various systems and methods for the analysis of gestural data are being developed. Purpose: A detailed review of methods and a comparative analysis of current approaches in automatic recognition of gestures and sign languages. Results: The main gesture recognition problems are the following: detection of articulators (mainly hands), pose estimation and segmentation of gestures in the flow of speech. The authors conclude that the use of two-stream convolutional and recurrent neural network architectures is generally promising for efficient extraction and processing of spatial and temporal features, thus solving the problem of dynamic gestures and coarticulations. This solution, however, heavily depends on the quality and availability of data sets. Practical relevance: This review can be considered a contribution to the study of rapidly developing sign language recognition, irrespective to particular natural sign languages. The results of the work can be used in the development of software systems for automatic gesture and sign language recognition.

Cumulative cultural evolution, population structure and the origin of combinatoriality in human language

Language is the primary repository and mediator of human collective knowledge. A central question for evolutionary linguistics is the origin of the combinatorial structure of language (sometimes referred to as duality of patterning), one of language’s basic design features. Emerging sign languages provide a promising arena to study the emergence of language properties. Many, but not all such sign languages exhibit combinatoriality, which generates testable hypotheses about its source. We hypothesize that combinatoriality is the inevitable result of learning biases in cultural transmission, and that population structure explains differences across languages. We construct an agent-based model with population turnover. Bayesian learning agents with a prior preference for compressible languages (modelling a pressure for language learnability) communicate in pairs under pressure to reduce ambiguity. We include two transmission conditions: agents learn the language either from the oldest agent or from an agent in the middle of their lifespan. Results suggest that (1) combinatoriality emerges during iterated cultural transmission under concurrent pressures for simplicity and expressivity and (2) population dynamics affect the rate of evolution, which is faster when agents learn from other learners than when they learn from old individuals. This may explain its absence in some emerging sign languages. We discuss the consequences of this finding for cultural evolution, highlighting the interplay of population-level, functional and cognitive factors. This article is part of a discussion meeting issue ‘The emergence of collective knowledge and cumulative culture in animals, humans and machines’.

New Perspectives on the Neurobiology of Sign Languages

The first 40 years of research on the neurobiology of sign languages (1960–2000) established that the same key left hemisphere brain regions support both signed and spoken languages, based primarily on evidence from signers with brain injury and at the end of the 20th century, based on evidence from emerging functional neuroimaging technologies (positron emission tomography and fMRI). Building on this earlier work, this review focuses on what we have learned about the neurobiology of sign languages in the last 15–20 years, what controversies remain unresolved, and directions for future research. Production and comprehension processes are addressed separately in order to capture whether and how output and input differences between sign and speech impact the neural substrates supporting language. In addition, the review includes aspects of language that are unique to sign languages, such as pervasive lexical iconicity, fingerspelling, linguistic facial expressions, and depictive classifier constructions. Summary sketches of the neural networks supporting sign language production and comprehension are provided with the hope that these will inspire future research as we begin to develop a more complete neurobiological model of sign language processing.

Object and handling handshapes in 11 sign languages: towards a typology of the iconic use of the hands

This article looks at cross-linguistic variation in lexical iconicity, addressing the question of to what extent and how this variation is patterned. More than in spoken languages, iconicity is highly frequent in the lexicons of sign languages. It is also highly complex, in that often multiple motivated components jointly shape an iconic lexeme. Recent typological research on spoken languages finds tentative iconic patterning in a large number of basic lexical items, underlining once again the significance of iconicity for human language. The uncontested and widespread use of iconicity found in the lexicons of sign languages enables us to take typological research into lexical iconicity to the next level. Indeed, previous studies have shown cross-linguistic variation in: a) the use of embodying and handling handshapes in sign languages (mostly of European origin) and b) the frequency of space-based size depiction in African and European sign languages. The two types of variation may be interrelated, as handling handshapes may use space-based size depiction. In this study, we first replicate earlier studies on the distribution of embodying and handling handshapes, this time in a data set consisting of a relatively large set of sign languages (n = 11), most of which are used in Africa. The results confirm significant variation across these sign languages. These findings are then compared to the use of space-based size depiction, revealing that these patterns independently from the distribution of embodying/handling handshapes. We argue that the results call for expanding typological studies on representational strategies in iconic signs beyond the now relatively well studied instrument/manipulation alternation. Fine-grained analyses on a multitude of iconic features in signs are likely to reveal cross-linguistic variation in iconic tendencies in SL lexicons.

Deaf Ni-Vanuatu and their signs: A sociolinguistic study

<p>This thesis provides the first documentation and description of the signs created and used by deaf individuals in Vanuatu. The specific aims of this research were as follows: to establish the sociolinguistic context experienced by deaf people in Vanuatu; to identify the repertoire and characteristics of signs used by the deaf participants; to compare features of participants’ individual signs with the characteristics of home signs and emerging sign languages; and to consider the degree of similarity and potential similarity of signs between participants and how this reflects individuals’ opportunities for contact with other deaf people and signing interlocutors. The limitations of this study are that field methodology for data collection was developed in situ as conditions allowed. The sociolinguistic context for deaf Ni-Vanuatu confirms that language isolation leads to marginalisation from community and society. The study established that these home sign lexicons were limited in quantity and conceptual range, and that shared background knowledge was essential for comprehension. Overall, 22 handshapes were documented, and the predominant handshapes unmarked. Most participants preferred handling strategy for depicting signs. Some evidence of noun-verb distinction was noted in the repertoire of some participants. However, across this range of formational characteristics, results showed significant individual variations. Furthermore, multiple barriers have precluded development of a shared sign language and any form of deaf community.</p>

Deaf Ni-Vanuatu and their signs: A sociolinguistic study

<p>This thesis provides the first documentation and description of the signs created and used by deaf individuals in Vanuatu. The specific aims of this research were as follows: to establish the sociolinguistic context experienced by deaf people in Vanuatu; to identify the repertoire and characteristics of signs used by the deaf participants; to compare features of participants’ individual signs with the characteristics of home signs and emerging sign languages; and to consider the degree of similarity and potential similarity of signs between participants and how this reflects individuals’ opportunities for contact with other deaf people and signing interlocutors. The limitations of this study are that field methodology for data collection was developed in situ as conditions allowed. The sociolinguistic context for deaf Ni-Vanuatu confirms that language isolation leads to marginalisation from community and society. The study established that these home sign lexicons were limited in quantity and conceptual range, and that shared background knowledge was essential for comprehension. Overall, 22 handshapes were documented, and the predominant handshapes unmarked. Most participants preferred handling strategy for depicting signs. Some evidence of noun-verb distinction was noted in the repertoire of some participants. However, across this range of formational characteristics, results showed significant individual variations. Furthermore, multiple barriers have precluded development of a shared sign language and any form of deaf community.</p>