Publications

Abstract

Proficient readers typically fixate near the center of a word, with a slight bias towards word onset. We explore a novel account of this phenomenon based on combining information-theory with visual perceptual constraints in a connectionist model of visual word recognition. This account posits that the amount of information-content available for word identification varies across fixation locations and across languages, thereby explaining the overall fixation location bias in different languages, making the novel prediction that certain words are more readily identified when fixating at an atypical fixation location, and predicting specific cross-linguistic differences. We tested these predictions across several simulations in English and Hebrew, and in a pilot behavioral experiment. Results confirmed that the bias to fixate closer to word onset aligns with maximizing information in the visual signal, that some words are more readily identified at atypical fixation locations, and that these effects vary to some degree across languages.

Abstract

Infants learn the meaning of words from accumulated experiences of real-time interactions with their caregivers. To study the effects of visual sensory input on word learning, we recorded infant's view of the world using head-mounted eye trackers during free-flowing play with a caregiver. While playing, infants were exposed to novel label-object mappings and later learning outcomes for these items were tested after the play session. In this study we use a classification based approach to link properties of infants' visual scenes during naturalistic labeling moments to their word learning outcomes. We find that a model which integrates both highly informative and ambiguous sensory evidence is a better fit to infants' individual learning outcomes than models where either type of evidence is taken alone, and that raw labeling frequency is unable to account for the word learning differences we observe. Here we demonstrate how a computational model, using only raw pixels taken from the egocentric scene image, can derive insights on human language learning.

Abstract

The RELISH project promotes language-oriented research by addressing a two-pronged problem: (1) the lack of harmonization between digital standards for lexical information in Europe and America, and (2) the lack of interoperability among existing lexicons of endangered languages, in particular those created with the Shoebox/Toolbox lexicon building software. The cooperation partners in the RELISH project are the University of Frankfurt (FRA), the Max Planck Institute for Psycholinguistics (MPI Nijmegen), and Eastern Michigan University, the host of the Linguist List (ILIT). The project aims at harmonizing key European and American digital standards whose divergence has hitherto impeded international collaboration on language technology for resource creation and analysis, as well as web services for archive access. Focusing on several lexicons of endangered languages, the project will establish a unified way of referencing lexicon structure and linguistic concepts, and develop a procedure for migrating these heterogeneous lexicons to a standards-compliant format. Once developed, the procedure will be generalizable to the large store of lexical resources involved in the LEGO and DoBeS projects.

Abstract

Data augmentation techniques have been widely used in visual recognition tasks as it is easy to generate new
data by simple and straight forward image transformations. However, when it comes to text data augmen-
tations, it is difficult to find appropriate transformation techniques which also preserve the contextual and
grammatical structure of language texts. In this paper, we explore various text data augmentation techniques
in text space and word embedding space. We study the effect of various augmented datasets on the efficiency
of different deep learning models for relation classification in text.

Abstract

Analysis of the functions of nominal apposition in a number of Latin authors representing different periods, genres, and
linguistic registers shows (1) that nominal apposition in Latin had a wide variety of functions; (2) that genre had some
effect on functional use; (3) that change did not affect semantic fields as such; and (4) that with time the occurrence of
apposition increasingly came to depend on the semantic field and within the semantic field on the individual lexical items.
The ‘per-word’ treatment –also attested for the structural development of nominal apposition– underscores the specific
characteristics of nominal apposition as a phenomenon at the cross-roads of syntax and derivational morphology

Abstract

The event-related potential (ERP) component named phonological mismatch negativity (PMN) arises when listeners hear an unexpected word form in a spoken sentence [1]. The PMN is thought to reflect the mismatch between expected and perceived auditory speech input. In this paper, we use the PMN to test a central premise in the predictive coding framework [2], namely that the mismatch between prior expectations and sensory input is an important mechanism of perception. We test this with natural speech materials containing approximately 50,000 word tokens. The corresponding EEG-signal was recorded while participants (n = 48) listened to these materials. Following [3], we quantify the mismatch with two word probability distributions (WPD): a WPD based on preceding context, and a WPD that is additionally updated based on the incoming audio of the current word. We use the between-WPD cross entropy for each word in the utterances and show that a higher cross entropy correlates with a more negative PMN. Our results show that listeners anticipate auditory input while processing each word in naturalistic speech. Moreover, complementing previous research, we show that predictive language processing occurs across the whole probability spectrum.

Abstract

The mismatch between top-down predicted and bottom-up perceptual input is an important mechanism of perception according to the predictive coding framework (Friston, [1]). In this paper we develop and validate a new information-theoretic measure that quantifies the mismatch between expected and observed auditory input during speech processing. We argue that such a mismatch measure is useful for the study of speech processing. To compute the mismatch measure, we use naturalistic speech materials containing approximately 50,000 word tokens. For each word token we first estimate the prior word probability distribution with the aid of statistical language modelling, and next use automatic speech recognition to update this word probability distribution based on the unfolding speech signal. We validate the mismatch measure with multiple analyses, and show that the auditory-based update improves the probability of the correct word and lowers the uncertainty of the word probability distribution. Based on these results, we argue that it is possible to explicitly estimate the mismatch between predicted and perceived speech input with the cross entropy between word expectations computed before and after an auditory update.

Abstract

The study of first language acquisition still strongly relies on behavioural methods to measure underlying linguistic abilities. In the present paper, we closely examine and model one such method, the headturn preference procedure (HPP), which is widely used to measure infant speech segmentation and word recognition abilities Our model takes real speech as input, and only uses basic sensory processing and cognitive capabilities to simulate observable behaviour.We show that the familiarity effect found in many HPP experiments can be simulated without using the phonetic and phonological skills necessary for segmenting test sentences into words. The explicit modelling of the process that converts the result of the cognitive processing of the test sentences into observable behaviour uncovered two issues that can lead to null-results in HPP studies. Our simulations show that caution is needed in making inferences about underlying language skills from behaviour in HPP experiments. The simulations also generated questions that must be addressed in future HPP studies.

Abstract

Studies have shown that cross-linguistically, consonants at the left edge of higher-level prosodic boundaries tend to be more forcefully articulated than those at lower-level boundaries, a phenomenon known as domain-initial strengthening. This study tests whether similar effects occur in Turkish, using the Autosegmental-Metrical model proposed by Ipek & Jun [1, 2] as the basis for assessing boundary strength. Productions of /t/ and /d/ were elicited in four domain-initial prosodic positions corresponding to progressively higher-level boundaries: syllable, word, intermediate phrase, and Intonational Phrase. A fifth position, nuclear word, was included in order to better situate it within the prosodic hierarchy. Acoustic correlates of articulatory strength were measured, including closure duration for /d/ and /t/, as well as voice onset time and burst energy for /t/. Our results show that closure duration increases cumulatively from syllable to intermediate phrase, while voice onset time and burst energy are not influenced by boundary strength. These findings provide corroborating evidence for Ipek & Jun’s model, particularly for the distinction between word and intermediate phrase boundaries. Additionally, articulatory strength at the left edge of the nuclear word patterned closely with word-initial position, supporting the view that the nuclear word is not associated with a distinct phrasing domain

Abstract

Although the possibility of referring or citing on-line data from publications is seen at least theoretically as an important means to provide immediate testable proof or simple illustration of a line of reasoning, the practice has not been wide-spread yet and no extensive experience has been gained about the possibilities and problems of referring to raw data-sets. This paper makes a case to investigate the possibility and need of persistent data visualization services that facilitate the inspection and evaluation of the cited data.

Abstract

This paper describes the status of the standardization efforts of a Component Metadata approach for describing Language Resources with metadata. Different linguistic and Language & Technology communities as CLARIN, META-SHARE and NaLiDa use this component approach and see its standardization of as a matter for cooperation that has the possibility to create a large interoperable domain of joint metadata. Starting with an overview of the component metadata approach together with the related semantic interoperability tools and services as the ISOcat data category registry and the relation registry we explain the standardization plan and efforts for component metadata within ISO TC37/SC4. Finally, we present information about uptake and plans of the use of component metadata within the three mentioned linguistic and L&T communities.

Abstract

This study investigates the lexical representation of second-language words that contain difficult to distinguish phonemes. Dutch and English listeners' perception of partially onset-overlapping word pairs like DAFFOdil-DEFIcit and minimal pairs like flash-flesh, was assessed with two cross-modal priming experiments, examining two stages of lexical processing: activation of intended and mismatching lexical representations (Exp.1) and competition between those lexical representations (Exp.2). Exp.1 shows that truncated primes like daffo- and defi- activated lexical representations of mismatching words (either deficit or daffodil) more for L2 than L1 listeners. Exp.2 shows that for minimal pairs, matching primes (prime: flash, target: FLASH) facilitated recognition of visual targets for L1 and L2 listeners alike, whereas mismatching primes (flesh, FLASH) inhibited recognition consistently for L1 listeners but only in a minority of cases for L2 listeners; in most cases, for them, primes facilitated recognition of both words equally strongly. Importantly, all listeners experienced a combination of facilitation and inhibition (and all items sometimes caused facilitation and sometimes inhibition). These results suggest that for all participants, some of the minimal pairs were represented with separate, native-like lexical representations, whereas other pairs were stored as homophones. The nature of the L2 lexical representations thus varied strongly even within listeners.

Abstract

Speech input causes listeners to activate multiple
candidate words which then compete with one
another. These include onset competitors, that share a
beginning (bumper, butter), but also, counterintuitively,
rhyme competitors, sharing an ending
(bumper, jumper). In L1, competition is typically
stronger for onset than for rhyme. In L2, onset
competition has been attested but rhyme competition
has heretofore remained largely unexamined. We
assessed L1 (Dutch) and L2 (English) word
recognition by the same late-bilingual individuals. In
each language, eye gaze was recorded as listeners
heard sentences and viewed sets of drawings: three
unrelated, one depicting an onset or rhyme competitor
of a word in the input. Activation patterns revealed
substantial onset competition but no significant
rhyme competition in either L1 or L2. Rhyme
competition may thus be a “luxury” feature of
maximally efficient listening, to be abandoned when
resources are scarcer, as in listening by late
bilinguals, in either language.

Abstract

Conversational turns often proceed with very brief pauses between speakers. In order to maintain “no gap, no overlap” turntaking, we must be able to anticipate when an ongoing utterance will end, tracking the current speaker for upcoming points of potential floor exchange. The precise set of cues that listeners use for turn-end boundary anticipation is not yet established. We used an eyetracking paradigm to measure adults’ and children’s online turn processing as they watched videos of conversations in their native language (English) and a range of other languages they did not speak. Both adults and children anticipated speaker transitions effectively. In addition, we observed evidence of turn-boundary anticipation for questions even in languages that were unknown to participants, suggesting that listeners’ success in turn-end anticipation does not rely solely on lexical information.

Abstract

Spontaneous gestures play an important role in spatial problem solving. We investigated the functional role and underlying mechanism of spontaneous gestures in spatial problem solving. In Experiment 1, 132 participants were required to solve a mental rotation task (see Figure 1) without speaking. Participants gestured more frequently in difficult trials than in easy trials. In Experiment 2, 66 new participants were given two identical sets of mental rotation tasks problems, as the one used in experiment 1. Participants who were encouraged to gesture in the first set of mental rotation task problemssolved more problems correctly than those who were allowed to gesture or those who were prohibited from gesturing both in the first set and in the second set in which all participants were prohibited from gesturing. The gestures produced by the gestureencouraged group and the gesture-allowed group were not qualitatively different. In Experiment 3, 32 new participants were first given a set of mental rotation problems and then a second set of nongesturing paper folding problems. The gesture-encouraged group solved more problems correctly in the first set of mental rotation problems and the second set of non-gesturing paper folding problems. We concluded that gesture improves spatial problem solving. Furthermore, gesture has a lasting beneficial effect even when gesture is not available and the beneficial effect is problem-general.We suggested that gesture enhances spatial problem solving by provide a rich sensori-motor representation of the physical world and pick up information that is less readily available to visuo-spatial processes.

Abstract

The nature of the connection between musical and spatial processing is controversial. While pitch may be described in spatial terms such as “high” or “low”, it is unclear whether pitch and space are associated but separate dimensions or whether they share representational and processing resources. In the present study, we asked participants to judge whether a target vocal note was the same as (or different from) a preceding cue note. Importantly, target trials were presented as video clips where a singer sometimes gestured upward or downward while singing that target note, thus providing an alternative, concurrent source of spatial information. Our results show that pitch discrimination was significantly biased by the spatial movement in gesture. These effects were eliminated by spatial memory load but preserved under verbal memory load conditions. Together, our findings suggest that pitch and space have a shared representation such that the mental representation of pitch is audiospatial in nature.

Abstract

In this study, we compared the performance of a long short-term memory (LSTM) neural network to the behavior of human participants on a language task that requires hierarchically structured knowledge. We show that humans interpret ambiguous noun phrases, such as second blue ball, in line with their hierarchical constituent structure. LSTMs, instead, only do
so after unambiguous training, and they do not systematically generalize to novel items. Overall, the results of our simulations indicate that a model can behave hierarchically without relying on hierarchical constituent structure.

Abstract

publication expected April 2012

Abstract

This study is a comprehensive description of the nominal demonstratives in Dalabon, a severely endangered Gunwinyguan non-Pama-Nyungan language of southwestern Arnhem Land, northern Australia. Demonstratives are attested in the basic vocabulary of every language, yet remain heretofore underdescribed in Australian languages. Traditional definitions of demonstratives as primarily making spatial reference have recently evolved at a great pace, with close analyses of demonstratives-in-use revealing that their use in spatial reference, in narrative discourse, and in interaction is significantly more complex than previously assumed, and that definitions of demonstrative forms are best developed after consideration of their use across these contexts. The present study reinforces findings of complexity in demonstrative use, and the significance of a multidimensional characterization of demonstrative forms. This study is therefore a contribution to the description of Dalabon, to the analysis of demonstratives in Australian languages, and to the theory and typology of demonstratives cross-linguistically. In this study, I present a multi-dimensional analysis of Dalabon demonstratives, using a variety of theoretical frameworks and research tools including descriptive linguistics, lexical-functional grammar, discourse analysis, gesture studies and pragmatics. Using data from personal narratives, improvised interactions and elicitation sessions to investigate the demonstratives, this study takes into account their morphosyntactic distribution, uses in the speech situation, interactional factors, discourse phenomena, concurrent gesture, and uses in personal narratives. I conclude with a unified account of the intenstional and extensional semantics of each form surveyed. The Dalabon demonstrative paradigm divides into two types, those which are spatially-specific and those which are non-spatial. The spatially-specific demonstratives nunda ‘this (in the here-space)’ and djakih ‘that (in the there-space)’ are shown not to encode the location of the referent per se, rather its relative position to dynamic physical and social elements of the speech situation such as the speaker’s engagement area and here-space. Both forms are also used as spatial adverbs to mean ‘here’ and ‘there’ respectively, while only nunda is also used as a temporal adverb ‘now, today’. The spatially-specific demonstratives are limited to situational use in narratives. The non-spatial demonstratives kanh/kanunh ‘that (identifiable)’ and nunh ‘that (unfamiliar, contrastive)’ are used in both the speech situation and personal narratives to index referents as ‘identifiable’ or ‘unfamiliar’ respectively. Their use in the speech situation can conversationally implicate that the referent is distal. The non-spatial demonstratives display the greatest diversity of use in narratives, each specializing for certain uses, yet their wide distribution across discourse usage types can be described on account of their intensional semantics. The findings of greatest typological interest in this study are that speakers’ choice of demonstrative in the speech situation is influenced by multiple simultaneous deictic parameters (including gesture); that oppositions in the Dalabon demonstrative paradigm are not equal, nor exclusively semantic; that the form nunh ‘that (unfamiliar, contrastive)’ is used to index a referent as somewhat inaccessible or unexpected; that the ‘recognitional’ form kanh/kanunh is instead described as ‘identifiable’; and that speakers use demonstratives to index emotional deixis to a referent, or to their addressee.

Abstract

Part of the remarkable efficiency of listening is
accommodation to unfamiliar talkers’ specific
pronunciations by retuning of phonemic intercategory
boundaries. Such retuning occurs in second
(L2) as well as first language (L1); however, recent
research with emigrés revealed successful adaptation
in the environmental L2 but, unprecedentedly, not in
L1 despite continuing L1 use. A possible explanation
involving relative exposure to novel talkers is here
tested in heritage language users with Mandarin as
family L1 and English as environmental language. In
English, exposure to an ambiguous sound in
disambiguating word contexts prompted the expected
adjustment of phonemic boundaries in subsequent
categorisation. However, no adjustment occurred in
Mandarin, again despite regular use. Participants
reported highly asymmetric interlocutor counts in the
two languages. We conclude that successful retuning
ability requires regular exposure to novel talkers in
the language in question, a criterion not met for the
emigrés’ or for these heritage users’ L1.

Abstract

A distinction between strong and weak vowels can be drawn on the basis of vowel quality, of stress, or of both factors. An experiment was conducted in which sets of contextually matched word-intial vowels ranging from clearly strong to clearly weak were cross-spliced, and the naturalness of the resulting words was rated by listeners. The ratings showed that in general cross-spliced words were only significantly less acceptable than unspliced words when schwa was not involved; this supports a categorical distinction based on vowel quality.

Abstract

Two lexical decision experiments in Cantonese are described in which the recognition of spoken target words as a function of phonological similarity to a preceding prime is investigated. Phonological similaritv in first syllables produced inhibition, while similarity in second syllables led to facilitation. Differences between syllables in tonal and segmental structure had generally similar effects.

Abstract

Speakers and listeners have a shared goal: to communicate. The processes of speech perception and of speech production interact in many ways under the constraints of this communicative goal; such interaction is as characteristic of prosodic processing as of the processing of other aspects of linguistic structure. Two of the major uses of prosodic information in situations of communication are to encode salience and segmentation, and these themes unite the contributions to the symposium introduced by the present review.

Abstract

People automatically chunk ongoing dynamic events into discrete units. This paper investigates whether linguistic structure is a factor in this process. We test the claim that describing an event with a serial verb construction will influence a speaker’s conceptual event structure. The grammar of Avatime (a Kwa language spoken in Ghana)requires its speakers to describe some, but not all, placement events using a serial verb construction which also encodes the preceding taking event. We tested Avatime and English speakers’ recognition memory for putting and taking events. Avatime speakers were more likely to falsely recognize putting and taking events from episodes associated with takeput serial verb constructions than from episodes associated with other constructions. English speakers showed no difference in false recognitions between episode types. This demonstrates that memory for episodes is related to the type of language used; and, moreover, across languages different conceptual representations are formed for the same physical episode, paralleling habitual linguistic practices

Abstract

Do interlocutors adjust their conversational strategies to the specific contextual demands of a given situation? Prior studies have yielded conflicting results, making it unclear how strategies vary with demands. We combine insights from qualitative and quantitative approaches in a within-participant experimental design involving two different contexts: spontaneously occurring conversations (SOC) and task-oriented conversations (TOC). We systematically assess backchanneling, other-repair and linguistic alignment. We find that SOC exhibit a higher number of backchannels, a reduced and more generic repair format and higher rates of lexical and syntactic alignment. TOC are characterized by a high number of specific repairs and a lower rate of lexical and syntactic alignment. However, when alignment occurs, more linguistic forms are aligned. The findings show that conversational strategies adapt to specific contextual demands.

Abstract

The present dissertation investigates what neurocognitive processes are activated in the brain when listeners comprehend spoken language and in particular the melody and rhythm of speech, also referred to as prosody. The findings of several electrophysiological studies show that prosody influences the early and late stages of spoken language processing. When words are accented, listeners consider them important, and the brain responds to accentuation already 200 milliseconds after stimulus onset. The processing of prosodic prominence occurs whether or not a context is present and whether or not accent is congruent with context, although the responses to accentuation may be modified by either of these factors and by the focus particle only. Listeners are sensitive not only to the presence of prosodic prominence but also to the type of accents speakers use: corrective prosody activates additional interpretation mechanisms related to the construction of corrective meaning. The parallel between accents across clauses impacts the disambiguation of sentences with verb ellipsis. By interpreting prosodically parallel elements as syntactically parallel, listeners arrive at less preferred interpretations of conjoined clauses. The research indentifies early correlates of incongruous prosody in strongly predictive contexts as well as late integration processes for prosody comprehension, which are related to the processing of structural complexity in isolated and ambiguous sentences. The dissertation provides evidence that the brain is sensitive to differences in prosody even in the absence of prosodic judgment. However, by changing the task, one modulates the neural mechanisms of prosody processing.

Abstract

We present a new transcription mode for the annotation tool ELAN. This mode is designed to speed up the process of creating transcriptions of primary linguistic data (video and/or audio recordings of linguistic behaviour). We survey the basic transcription workflow of some commonly used tools (Transcriber, BlitzScribe, and ELAN) and describe how the new transcription interface improves on these existing implementations. We describe the design of the transcription interface and explore some further possibilities for improvement in the areas of segmentation and computational enrichment of annotations.

Abstract

The widespread occurrence of ideophones, large classes of words specialized in evoking sensory imagery, is little known outside linguistics and anthropology. Ideophones are a common feature in many of the world’s languages but are underdeveloped in English and other Indo-European languages. Here we study the meanings of ideophones in Siwu (a Kwa language from Ghana) using a pile-sorting task. The goal was to uncover the underlying structure of the lexical space and to examine the claimed link between ideophones and perception. We found that Siwu ideophones are principally organized around fine-grained aspects of sensory perception, and map onto salient psychophysical dimensions identified in sensory science. The results ratify ideophones as dedicated sensory vocabulary and underline the relevance of ideophones for research on language and perception.

Abstract

People often talk about musical pitch in terms of spatial metaphors. In English, for instance, pitches can be high or low, whereas in other languages pitches are described as thick or thin. According to psychophysical studies, metaphors in language can also shape people’s nonlinguistic space-pitch representations. But does language establish mappings between space and pitch in the first place or does it modify preexisting associations? Here we tested 4-month-old Dutch infants’ sensitivity to height-pitch and thickness-pitch mappings in two preferential looking tasks. Dutch infants looked significantly longer at cross-modally congruent stimuli in both experiments, indicating that infants are sensitive to space-pitch associations prior to language. This early presence of space-pitch mappings suggests that these associations do not originate from language. Rather, language may build upon pre-existing mappings and change them gradually via some form of competitive associative learning.

Abstract

Overview Since the early nineties, the on-going dramatic loss of the world’s linguistic diversity has gained attention, first by the linguists and increasingly also by the general public. As a response, the new field of language documentation emerged from around 2000 on, starting with the funding initiative ‘Dokumentation Bedrohter Sprachen’ (DoBeS, funded by the Volkswagen foundation, Germany), soon to be followed by others such as the ‘Endangered Languages Documentation Programme’ (ELDP, at SOAS, London), or, in the USA, ‘Electronic Meta-structure for Endangered Languages Documentation’ (EMELD, led by the LinguistList) and ‘Documenting Endangered Languages’ (DEL, by the NSF). From its very beginning, the new field focused on digital technologies not only for recording in audio and video, but also for annotation, lexical databases, corpus building and archiving, among others. This development not just coincides but is intrinsically interconnected with the increasing focus on digital data, technology and methods in all sciences, in particular in the humanities.

Abstract

This contribution presents “The Language Archive” (TLA), a new unit at the MPI for Psycholinguistics, discussing the current developments in management of scientific data, considering the need for new data research infrastructures. Although several initiatives worldwide in the realm of language resources aim at the integration, preservation and mobilization of research data, the state of such scientific data is still often problematic. Data are often not well organized and archived and not described by metadata ― even unique data such as field-work observational data on endangered languages is still mostly on perishable carriers. New data centres are needed that provide trusted, quality-reviewed, persistent services and suitable tools and that take legal and ethical issues seriously. The CLARIN initiative has established criteria for suitable centres. TLA is in a good position to be one of such centres. It is based on three essential pillars: (1) A data archive; (2) management, access and annotation tools; (3) archiving and software expertise for collaborative projects. The archive hosts mostly observational data on small languages worldwide and language acquisition data, but also data resulting from experiments

Abstract

Previous studies have shown that speakers align their speech to each other at multiple linguistic levels. This study investigates whether alignment is mostly the result of priming from the immediately preceding
speech materials, focussing on pitch and articulation rate (AR). Native Dutch speakers completed sentences, first by themselves (pre-test), then in alternation with Confederate 1 (Round 1), with Confederate 2 (Round 2), with Confederate 1 again
(Round 3), and lastly by themselves again (post-test). Results indicate that participants aligned to the confederates and that this alignment lasted during the post-test. The confederates’ directly preceding sentences were not good predictors for the participants’ pitch and AR. Overall, the results indicate that alignment is more of a global effect than a local priming effect.

Abstract

1. Introduction Speech conveys not only linguistic meaning but also paralinguistic information, such as features of the speaker’s social background, physiology, and emotional state. Linguistic and paralinguistic information is encoded in speech by using largely the same vocal apparatus and both are transmitted simultaneously in the acoustic signal, drawing on a limited set of acoustic cues. How this simultaneous encoding is achieved, how the different types of information are disentangled by the listener, and how much they interfere with one another is presently not well understood. Previous research has highlighted the importance of acoustic source and filter cues for emotion and linguistic encoding respectively, which may suggest that the two types of information are encoded independently of each other. However, those lines of investigation have been almost completely disconnected (Murray & Arnott, 1993).

Abstract

Language resource centers allow researchers to reliably deposit their structured data together with associated meta data and run services operating on this deposited data. We are looking into possibilities to create long-term persistency of both the deposited data and the services operating on this data. Challenges, both technical and non-technical, that need to be solved are the need to replicate more than just the data, proper identification of the digital objects in a distributed environment by making use of persistent identifiers and the set-up of a proper authentication and authorization domain including the management of the authorization information on the digital objects. We acknowledge the investment that most language resource centers have made in their current infrastructure. Therefore one of the most important requirements is the loose coupling with existing infrastructures without the need to make many changes. This shift from a single language resource center into a federated environment of many language resource centers is discussed in the context of a real world center: The Language Archive supported by the Max Planck Institute for Psycholinguistics.

Abstract

Anaphora, a ubiquitous feature of natural language, poses a particular challenge to young children as they first learn language due to its referential ambiguity. In spite of this, parents and caregivers use anaphora frequently in child-directed speech, potentially presenting a risk to effective communication if children do not yet have the linguistic capabilities of resolving anaphora successfully. Through an eye-tracking study in a naturalistic free-play context, we examine the strategies that parents employ to calibrate their use of anaphora to their child's linguistic development level. We show that, in this way, parents are able to intuitively scaffold the complexity of their speech such that greater referential ambiguity does not hurt overall communication success.

Abstract

This paper shows that the dictation task, a well-
known testing instrument in language education, has
untapped potential as a research tool for studying
speech perception. We describe how transcriptions
can be scored on measures of lexical, orthographic,
phonological, and semantic similarity to target
phrases to provide comprehensive information about
accuracy at different processing levels. The former
three measures are automatically extractable,
increasing objectivity, and the middle two are
gradient, providing finer-grained information than
traditionally used. We evaluate the measures in an
English dictation task featuring phonetically reduced
continuous speech. Whereas the lexical and
orthographic measures emphasize listeners’ word
identification difficulties, the phonological measure
demonstrates that listeners can often still recover
phonological features, and the semantic measure
captures their ability to get the gist of the utterances.
Correlational analyses and a discussion of practical
and theoretical considerations show that combining
multiple measures improves the dictation task’s
utility as a research tool.

Abstract

Lexically guided perceptual learning has traditionally been studied with ambiguous consonant sounds to which native listeners are exposed in a purely receptive listening context. To extend previous research, we investigate whether lexically guided learning applies to a vowel shift encountered by non-native listeners in an interactive dialogue. Dutch participants played a two-player game in English in either a control condition, which contained no evidence for a vowel shift, or a lexically constraining condition, in which onscreen lexical information required them to re-interpret their interlocutor’s /ɪ/ pronunciations as representing /ε/. A phonetic categorization pre-test and post-test were used to assess whether the game shifted listeners’ phonemic boundaries such that more of the /ε/-/ɪ/ continuum came to be perceived as /ε/. Both listener groups showed an overall post-test shift toward /ɪ/, suggesting that vowel perception may be sensitive to directional biases related to properties of the speaker’s vowel space. Importantly, listeners in the lexically constraining condition made relatively more post-test /ε/ responses than the control group, thereby exhibiting an effect of lexically guided adaptation. The results thus demonstrate that non-native listeners can adjust their phonemic boundaries on the basis of lexical information to accommodate a vowel shift learned in interactive conversation.

Abstract

What is the key cognitive ability that makes humans unique among all the other animals? Our work aims at contributing to this research question adopting a comparative and philosophical approach to the origins of verbal language. In particular, we adopt three strands of analysis that are relevant in the context of comparative investigation on the the origins of verbal language: a) research on the evolutionary ‘homologies’, which provides information on the phylogenetic traits that humans and other primates share with their common ancestor; b) investigations on “analogous” traits, aimed at finding the evolutionary pressures that guided the emergence of the same biological traits that evolved independently in phylogenetically distant species; the ontogenetic development of the ability to produce and understand verbal language in human infants. Within this comparative approach, we focus on three key apsects that we addressed bridging recent empiric evidence on language processing with philosophical investigations on verbal language: (i) pattern processing as a biologocal precursor of syntax and algebraic rule acquisition, (ii) sound modulation as a guide to pattern comprehension in speech, animal vocalization and music, (iii) social strategies for mutual understanding, survival and group cohesion. We conclude emphasizing the interplay between these three sets of cognitive processes as a fundamental dimension grounding the emergence of the human ability for propositional language.

Abstract

As knowing a foreign language becomes a necessity in the modern world, a large portion of
the population is faced with the challenge of learning a language in a classroom. This, in turn,
presents a unique set of difficulties. Acquiring a language with limited and artificial exposure makes
learning new information and vocabulary particularly difficult. The purpose of this thesis is to help us
understand how we can compensate—at least partially—for these difficulties by presenting
information in a way that aids learning. In particular, I focused on variables that affect semantic
richness—meaning the amount and variability of information associated with a word. Some factors
that affect semantic richness are intrinsic to the word and others pertain to that word’s relationship
with other items and information. This latter group depends on the context around the to-be-
learned items rather than the words themselves. These variables are easier to manipulate than
intrinsic qualities, making them more accessible tools for teaching and understanding learning. I
focused on two factors: emotionality of the surrounding semantic context and contextual diversity.
Publication 1 (Frances, de Bruin, et al., 2020b) focused on content learning in a foreign
language and whether the emotionality—positive or neutral—of the semantic context surrounding
key information aided its learning. This built on prior research that showed a reduction in
emotionality in a foreign language. Participants were taught information embedded in either
positive or neutral semantic contexts in either their native or foreign language. When they were
then tested on these embedded facts, participants’ performance decreased in the foreign language.
But, more importantly, they remembered better the information from the positive than the neutral
semantic contexts.
In Publication 2 (Frances, de Bruin, et al., 2020a), I focused on how emotionality affected
vocabulary learning. I taught participants the names of novel items described either in positive or
neutral terms in either their native or foreign language. Participants were then asked to recall and
recognize the object's name—when cued with its image. The effects of language varied with the
difficulty of the task—appearing in recall but not recognition tasks. Most importantly, learning the
words in a positive context improved learning, particularly of the association between the image of
the object and its name.
In Publication 3 (Frances, Martin, et al., 2020), I explored the effects of contextual
diversity—namely, the number of texts a word appears in—on native and foreign language word
learning. Participants read several texts that had novel pseudowords. The total number of
encounters with the novel words was held constant, but they appeared in 1, 2, 4, or 8 texts in either
their native or foreign language. Increasing contextual diversity—i.e., the number of texts a word
appeared in—improved recall and recognition, as well as the ability to match the word with its
meaning. Using a foreign language only affected performance when participants had to quickly
identify the meaning of the word.
Overall, I found that the tested contextual factors related to semantic richness—i.e.,
emotionality of the semantic context and contextual diversity—can be manipulated to improve
learning in a foreign language. Using positive emotionality not only improved learning in the foreign
language, but it did so to the same extent as in the native language. On a theoretical level, this
suggests that the reduction in emotionality in a foreign language is not ubiquitous and might relate
to the way in which that language as learned.
The third article shows an experimental manipulation of contextual diversity and how this
can affect learning of a lexical item, even if the amount of information known about the item is kept
constant. As in the case of emotionality, the effects of contextual diversity were also the same
between languages. Although deducing words from context is dependent on vocabulary size, this
does not seem to hinder the benefits of contextual diversity in the foreign language.
Finally, as a whole, the articles contained in this compendium provide evidence that some
aspects of semantic richness can be manipulated contextually to improve learning and memory. In
addition, the effects of these factors seem to be independent of language status—meaning, native
or foreign—when learning new content. This suggests that learning in a foreign and a native
language is not as different as I initially hypothesized, allowing us to take advantage of native
language learning tools in the foreign language, as well.

Abstract

Achieving linguistic proficiency requires identifying words from speech, and discovering the constraints that govern the way those words are used. In a recent study of non-adjacent dependency learning, Frost and Monaghan (2016) demonstrated that learners may perform these tasks together, using similar statistical processes - contrary to prior suggestions. However, in their study, non-adjacent dependencies were marked by phonological cues (plosive-continuant-plosive structure), which may have influenced learning. Here, we test the necessity of these cues by comparing learning across three conditions; fixed phonology, which contains these cues, varied phonology, which omits them, and shapes, which uses visual shape sequences to assess the generality of statistical processing for these tasks. Participants segmented the sequences and generalized the structure in both auditory conditions, but learning was best when phonological cues were present. Learning was around chance on both tasks for the visual shapes group, indicating statistical processing may critically differ across domains.

Abstract

In many languages the future is in front and the past behind, but in some cultures (like Aymara) the past is in front. Is it possible to find this mapping as an alternative conceptualization of time in other cultures? If so, what are the factors that affect its choice out of the set of available alternatives? In a paper and pencil task, participants placed future or past events either in front or behind a character (a schematic head viewed from above). A sample of 24 Islamic participants (whose language also places the future in front and the past behind) tended to locate the past event in the front box more often than Spanish participants. This result might be due to the greater cultural value assigned to tradition in Islamic culture. The same pattern was found in a sample of Spanish elders (N = 58), what may support that conclusion. Alternatively, the crucial factor may be the amount of attention paid to the past. In a final study, young Spanish adults (N = 200) who had just answered a set of questions about their past showed the past-in-front pattern, whereas questions about their future exacerbated the future-in-front pattern. Thus, the attentional explanation was supported: attended events are mapped to front space in agreement with the experiential connection between attending and seeing. When attention is paid to the past, it tends to occupy the front location in spite of available alternative mappings in the language-culture.

Abstract

Caused motion events (e.g. a boy pulls a box into a room) are basic events where an Agent (the boy) performs an Action (pulling) that causes a Figure (box) to move in a spatial Path (into) to a Goal (the room). These semantic elements are mapped onto lexical and syntactic structures differently across languages This dissertation investigates the encoding of caused motion events in Turkish, and the development of this encoding in speech and gesture. First, a linguistic analysis shows that Turkish does not fully fit into the expected typological patterns, and that the encoding of caused motion is determined by the fine-grained lexical semantics of a verb as well as the syntactic construction the verb is integrated into. A grammaticality judgment study conducted with adult Turkish speakers further establishes the fundamentals of the encoding patterns. An event description study compares adults’ verbal and gestural representations of caused motion to those of children aged 3 to 5. The findings indicate that although language-specificity is evident in children’s speech and gestures, the development of adult patterns takes time and occurs after the age of 5. A final study investigates a longitudinal video corpus of the spontaneous speech of Turkish-speaking children aged 1 to 3, and finds that language-specificity is evident from the start in both children’s speech and gesture. Apart from contributing to the literature on the development of Turkish, this dissertation furthers our understanding of the interaction between language-specificity and the multimodal expression of semantic information in event descriptions.

Abstract

Graph neural networks (GNNs) have emerged as the standard method for numerous tasks on graph-structured data such as node classification. However, real-world graphs are often evolving over time and even new classes may arise. We model these challenges as an instance of lifelong learning, in which a learner faces a sequence of tasks and may take over knowledge acquired in past tasks. Such knowledge may be stored explicitly as historic data or implicitly within model parameters. In this work, we systematically analyze the influence of implicit and explicit knowledge. Therefore, we present an incremental training method for lifelong learning on graphs and introduce a new measure based on k-neighborhood time differences to address variances in the historic data. We apply our training method to five representative GNN architectures and evaluate them on three new lifelong node classification datasets. Our results show that no more than 50% of the GNN's receptive field is necessary to retain at least 95% accuracy compared to training over the complete history of the graph data. Furthermore, our experiments confirm that implicit knowledge becomes more important when fewer explicit knowledge is available.

Abstract

Large-scale graph data in real-world applications is often not static but dynamic,
i. e., new nodes and edges appear over time. Current graph convolution approaches
are promising, especially, when all the graph’s nodes and edges are available dur-
ing training. When unseen nodes and edges are inserted after training, it is not
yet evaluated whether up-training or re-training from scratch is preferable. We
construct an experimental setup, in which we insert previously unseen nodes and
edges after training and conduct a limited amount of inference epochs. In this
setup, we compare adapting pretrained graph neural networks against retraining
from scratch. Our results show that pretrained models yield high accuracy scores
on the unseen nodes and that pretraining is preferable over retraining from scratch.
Our experiments represent a first step to evaluate and develop truly online variants
of graph neural networks.

Abstract

COVID-19 research datasets are crucial for analyzing research dynamics. Most collections of COVID-19 research items do not to include cited works and do not have annotations
from a controlled vocabulary. Starting with ZB MED KE data on COVID-19, which comprises CORD-19, we assemble a new dataset that includes cited work and MeSH annotations for all records. Furthermore, we conduct experiments on the analysis of research dynamics, in which we investigate predicting links in a co-annotation graph created on the basis of the new dataset. Surprisingly, we find that simple heuristic methods are better at
predicting future links than more sophisticated approaches such as graph neural networks.

Abstract

Print Friendly XML Gebre, Binyam Gebrekidan, Max Planck Institute for Psycholinguistics, The Netherlands, binyamgebrekidan.gebre [at] mpi.nl Wittenburg, Peter, Max Planck Institute for Psycholinguistics, The Netherlands, peter.wittenburg [at] mpi.nl Introduction Many gesture and sign language researchers manually annotate video recordings to systematically categorize, analyze and explain their observations. The number and kinds of annotations are so diverse and unpredictable that any attempt at developing non-adaptive automatic annotation systems is usually less effective. The trend in the literature has been to develop models that work for average users and for average scenarios. This approach has three main disadvantages. First, it is impossible to know beforehand all the patterns that could be of interest to all researchers. Second, it is practically impossible to find enough training examples for all patterns. Third, it is currently impossible to learn a model that is robustly applicable across all video quality-recording variations.

Abstract

Automatic annotation of gesture strokes is important for many gesture and sign language researchers. The unpredictable diversity of human gestures and video recording conditions require that we adopt a more adaptive case-by-case annotation model. In this paper, we present a work-in progress annotation model that allows a user to a) track hands/face b) extract features c) distinguish strokes from non-strokes. The hands/face tracking is done with color matching algorithms and is initialized by the user. The initialization process is supported with immediate visual feedback. Sliders are also provided to support a user-friendly adjustment of skin color ranges. After successful initialization, features related to positions, orientations and speeds of tracked hands/face are extracted using unique identifiable features (corners) from a window of frames and are used for training a learning algorithm. Our preliminary results for stroke detection under non-ideal video conditions are promising and show the potential applicability of our methodology.

Abstract

Recognizing the speech acts in our interlocutors’ utterances is a crucial prerequisite for conversation. However, it is not a trivial task given that the form and content of utterances is frequently underspecified for this level of meaning. In the present study we investigate participants’ competence in categorizing speech acts in such action-underspecific sentences and explore the time-course of speech act inferencing using a self-paced reading paradigm. The results demonstrate that participants are able to categorize the speech acts with very high accuracy, based on limited context and without any prosodic information. Furthermore, the results show that the exact same sentence is processed differently depending on the speech act it performs, with reading times starting to differ already at the first word. These results indicate that participants are very good at “getting” the speech acts, opening up a new arena for experimental research on action recognition in conversation.

Abstract

“The most profound technologies are those that disappear. They weave themselves into the fabric of everyday life until they are indistinguishable from it.” (Weiser 1991, S. 94). – Die Behauptung stammt aus einem vielzitierten Text von Mark Weiser, ehemals Chief Technology Officer am berühmten Xerox Palo Alto Research Center (PARC), wo nicht nur einige bedeutende computertechnische Innovationen ihren Ursprung hatten, sondern auch grundlegende anthropologische Einsichten zum Umgang mit technischen Artefakten gewonnen wurden.1 In einem populärwissenschaftlichen Artikel mit dem Titel „The Computer for the 21st Century” entwarf Weiser 1991 die Vision einer Zukunft, in der wir nicht mehr mit einem einzelnen PC an unserem Arbeitsplatz umgehen – vielmehr seien wir in jedem Raum umgeben von hunderten elektronischer Vorrichtungen, die untrennbar in Alltagsgegenstände eingebettet und daher in unserer materiellen Umwelt gleichsam „verschwunden“ sind. Dabei ging es Weiser nicht allein um das ubiquitäre Phänomen, das in der Medientheorie als „Transparenz der Medien“ bekannt ist2 oder in allgemeineren Theorien der Alltagserfahrung als eine selbstverständliche Verwobenheit des Menschen mit den Dingen, die uns in ihrem Sinn vertraut und praktisch „zuhanden“ sind.3 Darüber hinaus zielte Weisers Vision darauf, unsere bereits existierende Umwelt durch computerlesbare Daten zu erweitern und in die Operationen eines solchen allgegenwärtigen Netzwerks alltägliche Praktiken gleichsam lückenlos zu integrieren: In der Welt, die Weiser entwirft, öffnen sich Türen für denjenigen, der ein bestimmtes elektronisches Abzeichen trägt, begrüßen Räume Personen, die sie betreten, mit Namen, passen sich Computerterminals an die Präferenzen individueller Nutzer an usw. (Weiser 1991, S. 99).

Abstract

The standard for the analysis of distorted voices is perceptual rating of read-out texts or spontaneous speech. Automatic voice evaluation, however, is usually done on stable sections of sustained vowels. In this paper, text-based and established vowel-based analysis are compared with respect to their ability to measure hoarseness and its subclasses. 73 hoarse patients (48.3±16.8 years) uttered the vowel /e/ and read the German version of the text “The North Wind and the Sun”. Five speech therapists and physicians rated roughness, breathiness, and hoarseness according to the German RBH evaluation scheme. The best human-machine correlations were obtained for measures based on the Cepstral Peak Prominence (CPP; up to |r | = 0.73). Support Vector Regression (SVR) on CPP-based measures and prosodic features improved the results further to r ≈0.8 and confirmed that automatic voice evaluation should be performed on a text recording.

Abstract

Child songs are a great source for linguistic learning. Here we explore whether children can acquire novel words in a second language by playing a game featuring child songs in a playhouse. We present data from three studies that serve as scientific proof for the functionality of one game of the playhouse: the Energy Center. For this game, three hand-bikes were mounted on a panel. When children start moving the hand-bikes, child songs start playing simultaneously. Once the children produce enough energy with the hand-bikes, the songs are additionally accompanied with the sounds of musical instruments. In our studies, children executed a picture-selection task to evaluate whether they acquired new vocabulary from the songs presented during the game. Two of our studies were run in the field, one at a Dutch and one at an Indian pre-school. The third study features data from a more controlled laboratory setting. Our results partly confirm that the Energy Center is a successful means to support vocabulary acquisition in a second language. More research with larger sample sizes and longer access to the Energy Center is needed to evaluate the overall functionality of the game. Based on informal observations at our test sites, however, we are certain that children do pick up linguistic content from the songs during play, as many of the children repeat words and phrases from songs they heard. We will pick up upon these promising observations during future studies

Abstract

This study investigated whether a shadowing task can provide insights in the nature of reduction processes that are typical of casual speech. We focused on the shortening and presence versus absence of schwa and /t/ in Dutch past participles. Results showed that the absence of these segments was affected by the same variables as their shortening, suggesting that absence mostly resulted from extreme gradient shortening. This contrasts with results based on recordings of spontaneous conversations. We hypothesize that this difference is due to non-casual fast speech elicited by a shadowing task.

Abstract

Children with developmental language disorder (DLD) regularly use the base form of verbs (e.g., dance) instead of inflected forms (e.g., danced). We propose an account of this behavior in which children with DLD have difficulty processing novel inflected verbs in their input. This leads the inflected form to face stronger competition from alternatives. Competition is resolved by the production of a more accessible alternative with high semantic overlap with the inflected form: in English, the bare form. We test our account computationally by training a nonparametric Bayesian model that infers the productivity of the inflectional suffix (-ed). We systematically vary the number of novel types of inflected verbs in the input to simulate the input as processed by children with and without DLD. Modeling results are consistent with our hypothesis, suggesting that children’s inconsistent use of inflectional morphemes could stem from inferences they make on the basis of impoverished data.

Abstract

Prediction in language has traditionally been studied using
simple designs in which neural responses to expected
and unexpected words are compared in a categorical
fashion. However, these designs have been contested
as being ‘prediction encouraging’, potentially exaggerating
the importance of prediction in language understanding.
A few recent studies have begun to address
these worries by using model-based approaches to probe
the effects of linguistic predictability in naturalistic stimuli
(e.g. continuous narrative). However, these studies
so far only looked at very local forms of prediction, using
models that take no more than the prior two words into
account when computing a word’s predictability. Here,
we extend this approach using a state-of-the-art neural
language model that can take roughly 500 times longer
linguistic contexts into account. Predictability estimates
fromthe neural network offer amuch better fit to EEG data
from subjects listening to naturalistic narrative than simpler
models, and reveal strong surprise responses akin to
the P200 and N400. These results show that predictability
effects in language are not a side-effect of simple designs,
and demonstrate the practical use of recent advances
in AI for the cognitive neuroscience of language.

Abstract

Using the visual-word paradigm, the present study investigated the effects of word onset and offset masking on the time course of non-native spoken-word recognition in the presence of background noise. In two experiments, Dutch non-native listeners heard English target words, preceded by carrier sentences that were noise-free (Experiment 1) or contained intermittent noise (Experiment 2). Target words were either onset- or offset-masked or not masked at all. Results showed that onset masking delayed target word recognition more than offset masking did, suggesting that – similar to natives – non-native listeners strongly rely on word onset information during word recognition in noise.

Abstract

Co-speech gestures are an integral part of human face-to-face communication, but little is known about how pragmatic factors influence our comprehension of those gestures. The present study investigates how different types of recipients process iconic gestures in a triadic communicative situation. Participants (N = 32) took on the role of one of two recipients in a triad and were presented with 160 video clips of an actor speaking, or speaking and gesturing. Crucially, the actor’s eye gaze was manipulated in that she alternated her gaze between the two recipients. Participants thus perceived some messages in the role of addressed recipient and some in the role of unaddressed recipient. In these roles, participants were asked to make judgements concerning the speaker’s messages. Their reaction times showed that unaddressed recipients did comprehend speaker’s gestures differently to addressees. The findings are discussed with respect to automatic and controlled processes involved in gesture comprehension.

Abstract

This study investigates effects of prosodic factors (prominence, boundary) on coarticulatory Vnasalization in Australian English (AusE) in CVN and NVC in comparison to those in American English
(AmE). As in AmE, prominence was found to
lengthen N, but to reduce V-nasalization, enhancing N’s nasality and V’s orality, respectively (paradigmatic contrast enhancement). But the prominence effect in CVN was more robust than that in AmE. Again similar to findings in AmE, boundary
induced a reduction of N-duration and V-nasalization phrase-initially (syntagmatic contrast enhancement), and increased the nasality of both C and V phrasefinally.
But AusE showed some differences in terms
of the magnitude of V nasalization and N duration. The results suggest that the linguistic contrast enhancements underlie prosodic-structure modulation of coarticulatory V-nasalization in
comparable ways across dialects, while the fine phonetic detail indicates that the phonetics-prosody interplay is internalized in the individual dialect’s phonetic grammar.

Abstract

There is a strong relation between children’s exposure to
spatial terms and their later memory accuracy. In the current
study, we tested whether the production of spatial terms by
children themselves predicts memory accuracy and whether
and how language modality of these encodings modulates
memory accuracy differently. Hearing child speakers of
Turkish and deaf child signers of Turkish Sign Language
described pictures of objects in various spatial relations to each
other and later tested for their memory accuracy of these
pictures in a surprise memory task. We found that having
described the spatial relation between the objects predicted
better memory accuracy. However, the modality of these
descriptions in sign, speech, or speech-plus-gesture did not
reveal differences in memory accuracy. We discuss the
implications of these findings for the relation between spatial
language, memory, and the modality of encoding.

Abstract

In different fields of the humanities annotations of multimodal resources are a necessary component of the research workflow. Examples include linguistics, psychology, anthropology, etc. However, creation of those annotations is a very laborious task, which can take 50 to 100 times the length of the annotated media, or more. This can be significantly improved by applying innovative audio and video processing algorithms, which analyze the recordings and provide automated annotations. This is the aim of the AVATecH project, which is a collaboration of the Max Planck Institute for Psycholinguistics (MPI) and the Fraunhofer institutes HHI and IAIS. In this paper we present a set of results of automated annotation together with an evaluation of their quality.

Abstract

Introduction Human spoken communication is multimodal, i.e. it encompasses both speech and gesture. Acoustic properties of voice, body movements, facial expression, etc. are an inherent and meaningful part of spoken interaction; they can provide attitudinal, grammatical and semantic information. In the recent years interest in audio-visual corpora has been rising rapidly as they enable investigation of different communicative modalities and provide more holistic view on communication (Kipp et al. 2009). Moreover, for some languages such corpora are the only available resource, as is the case for endangered languages for which no written resources exist.

Abstract

In this paper we describe audio and video processing algorithms that are developed in the scope of AVATecH project. The purpose of these algorithms is to shorten the time taken by manual annotation of audio and video recordings by extracting features from media files and creating semi-automated annotations. We show that the use of such supporting algorithms can shorten the annotation time to 30-50% of the time necessary to perform a fully manual annotation of the same kind.