Voice Type Research Articles

Robot Voices in Daily Life: Vocal Human-Likeness and Application Context as Determinants of User Acceptance.

The growing popularity of speech interfaces goes hand in hand with the creation of synthetic voices that sound ever more human. Previous research has been inconclusive about whether anthropomorphic design features of machines are more likely to be associated with positive user responses or, conversely, with uncanny experiences. To avoid detrimental effects of synthetic voice design, it is therefore crucial to explore what level of human realism human interactors prefer and whether their evaluations may vary across different domains of application. In a randomized laboratory experiment, 165 participants listened to one of five female-sounding robot voices, each with a different degree of human realism. We assessed how much participants anthropomorphized the voice (by subjective human-likeness ratings, a name-giving task and an imagination task), how pleasant and how eerie they found it, and to what extent they would accept its use in various domains. Additionally, participants completed Big Five personality measures and a tolerance of ambiguity scale. Our results indicate a positive relationship between human-likeness and user acceptance, with the most realistic sounding voice scoring highest in pleasantness and lowest in eeriness. Participants were also more likely to assign real human names to the voice (e.g., “Julia” instead of “T380”) if it sounded more realistic. In terms of application context, participants overall indicated lower acceptance of the use of speech interfaces in social domains (care, companionship) than in others (e.g., information & navigation), though the most human-like voice was rated significantly more acceptable in social applications than the remaining four. While most personality factors did not prove influential, openness to experience was found to moderate the relationship between voice type and user acceptance such that individuals with higher openness scores rated the most human-like voice even more positively. Study results are discussed in the light of the presented theory and in relation to open research questions in the field of synthetic voice design.

Music & Poetry

The following conversations with contemporary classical composers demonstrate how music and poetry speak much the same language, and even understand themselves in each other's terms regarding rhythm and sound, time and structure, address and listeners, recitals and recitation. Kate Soper, Matthew Aucoin, and Seth Parker Woods are three very different American composers working today. Soper also writes her own libretti—syncretic texts that engage classical and medieval verse and contemporary literature. An excerpt of her recent work, Romance of the Rose, is included here. Aucoin, an erstwhile poet, whose Eurydice was performed this season at the Metropolitan Opera, has set to music poems and other writings by Walt Whitman, Paul Celan, and Jorie Graham. Woods, a cellist and performance artist, often interweaves archival recordings and live poetry recitation on stage, while also engaging visual arts. The work of all three composers is set to be featured this year at the 76th Ojai Music Festival in Ojai, California. (American Modern Opera Company, or AMOC, co-founded by Aucoin, is the first collective to be chosen as the festival's music director.) These discussions consider the spoken word, voice types, and silence; lecture-performance, the academy, canons, and avant-gardes; the interplay between history, material texts, and graphic notation; in addition to the creative cycle of dreaming, composition, and decay.Geoffrey Lokke

Speaker Identification Based on Physical Variation of Speech Signal

Speaker identification for the speech signal processing request, determining the speaker is a challenge due to physical variation. This paper emphasizes a new algorithm based on acoustic feature analysis of text-dependent speech. In this proposed method text-dependent speech changed by ten physical variation methods. Acoustic feature of all types of voice is calculated by its arithmetical correlation coefficients and mean value. The audio characteristic is calculated with Mel-frequency cepstrum coefficient (MFCC), its derivatives and double derivatives. An acoustic characteristic is analysed by using normal voice and changed voice by different speakers, the mixed data used for test and training purpose. Passing all the training and test data through the various classifiers based on identification system. Speaker identification efficiency results are calculated from the different classifier.

Effects of Age and Parkinson's Disease on the Relationship between Vocal Fold Abductory Kinematics and Relative Fundamental Frequency

This study reports on two experiments to examine vocal fold abduction and its relationship with relative fundamental frequency (RFF), considering two attributes that have been shown to elicit group differences in RFF: age (Experiment 1) and Parkinson's disease (PD; Experiment 2). For both experiments, simultaneous acoustic and nasendoscopic recordings were collected as participants produced the utterance, /ifi/. RFF values were computed from the acoustic signal, whereas abduction duration and glottic angle at voicing offset were identified from the laryngoscopic images. In Experiment 1, 50 speakers with typical voices (18-83 years) were analyzed to examine (1A) the effects of speaker age on individual outcome measures (RFF, abduction duration, glottic angle) via Pearson's correlation coefficients, and (1B) the effects of abductory measures and age on RFF via an analysis of covariance. In Experiment 2, 20 speakers with PD and 20 matched controls were analyzed to examine (2A) the effects of group (with/without PD) on outcome measures via an analysis of variance, and (2B) the relationship of RFF with abduction duration, glottic angle, and age when considering group via an analysis of covariance. Age demonstrated a significant, negative relationship with glottic angle (1A) but was not a significant factor when examining the relationship of vocal fold abduction and RFF (1B). Speaker group (with/without PD) demonstrated a significant effect on measures of RFF and abduction duration (2A) but was not a significant factor when examining the relationship of vocal fold abduction and RFF (2B). RFF is sensitive to changes in vocal fold abductory patterns during devoicing, irrespective of speaker age or PD status.

The clear speech intelligibility benefit for text-to-speech voices: Effects of speaking style and visual guise

This study examined how speaking style and guise influence the intelligibility of text-to-speech (TTS) and naturally produced human voices. Results showed that TTS voices were less intelligible overall. Although using a clear speech style improved intelligibility for both human and TTS voices (using "newscaster" neural TTS), the clear speech effect was stronger for TTS voices. Finally, a visual device guise decreased intelligibility, regardless of voice type. The results suggest that both speaking style and visual guise affect intelligibility of human and TTS voices. Findings are discussed in terms of theories about the role of social information in speech perception.

Sympathy for the digital: Influence of synthetic voice on affinity, social presence and empathy for photorealistic virtual humans

In this paper, we investigate the effect of a realism mismatch in the voice and appearance of a photorealistic virtual character in both immersive and screen-mediated virtual contexts. While many studies have investigated voice attributes for robots, not much is known about the effect voice naturalness has on the perception of realistic virtual characters. We conducted the first experiment in Virtual Reality (VR) with over two hundred participants investigating the mismatch between realistic appearance and unrealistic voice on the feeling of presence, and the emotional response of the user to the character expressing a strong negative emotion. We predicted that the mismatched voice would lower social presence and cause users to have a negative emotional reaction and feelings of discomfort towards the character. We found that the concern for the virtual character was indeed altered by the unnatural voice, though interestingly it did not affect social presence. The second experiment was conducted with a view towards heightening the appearance realism of the same character for the same scenarios, with an additional lower level of voice realism employed to strengthen the mismatch of perceptual cues. While voice type did not appear to impact reports of empathic responses towards the character, there was an observed effect of voice realism on reported social presence, which was not detected in the first study. There were also significant results on affinity and voice trait measurements that provide evidence in support of perceptual mismatch theories of the Uncanny Valley.

Enough is enough: how much intonation is needed in the vocal delivery of audio description?

ABSTRACT In this article we report on an experiment examining audience experience when two different audio description (AD) intonation types are used. Following the classification put forward by Cabeza-Cáceres (2013), we tested twenty participants with vision loss who were asked to listen to twenty film clips with AD voiced following the adapted and emphatic intonation. All clips were rated as evoking emotions, either positive or negative. Participants’ experience was measured through heart rate variability, a self-assessment emotional response (Self-Assessment Manikin), a self-assessment presence questionnaire (ITC-SOPI) and an interview. Results show that participants strongly preferred adapted intonation, particularly when presented clips were evoking negative emotions. Higher rating of intonation type was also linked to greater intensity of the emotional experience and immersion, both self-reported and measured as psychophysiological reaction. The qualitative analysis of the participants’ reports on their experience indicated that adapted intonation is considered a golden mean between emotionless and melodramatic intonation. Film genre, voice type and emotional valence can be influencing factors. Emphatic intonation could be beneficial for certain genres, but a poorly-matched voice can distract the audiences.

Testing the Strategic Response Model to Show Joint Effect of Resource Dependency and Centrality in Donor Network on NGO Response to Donor Demand

The Strategic Response Model (SRM) integrates two constructs, an organization’s resource dependence and network centrality, to predict response to an external demand. This article puts the SRM to test to demonstrate its applicability as a management tool to help with decision-making. Using forty-nine Lebanese nongovernmental environmental organizations’ (NGOs’) responses to bilateral donors who changed funding interests, the results are consistent with the model’s prediction of three types of responses, exit, voice, and adjustment, regardless of which of three resource dependency variables are used. To add context to this test of the SRM model, the dynamics within a larger system of resource pursuit and allocation across sectors, especially for non-Western settings characterized by turbulence and uncertainty, are discussed. Donors and nonprofits need to consider short- and long-term strategic decisions, knowing that relationships created and fostered may be as important as resources provided and consumed.

VOCAL TONE PRECISION DETECTION USING HARMONIC PRODUCT SPECTRUM (HPS) AND K-NEAREST NEIGHBOR (KNN) CLASSIFICATION

The progress of the digital era that is happening today, encourages rapid development in technology and science, one of which is in the field of art. Of all performing arts, the art of singing is the most complex, which requires a lot of preparation and practice. Everyone has a different type of voice. Males generally have three types of voice, namely bass, baritone, and tenor, while women generally have three types of voice, namely contralto (alto), mezzo-soprano, and soprano. However, not everyone knows what kind of voice they have. Therefore, this study will focus on classifying the human voice. In this study, the author uses the Harmonic Product Spectrum (HPS) and K-Nearest Neighbor (K-NN) algorithms. The data used is in the form of primary voice recording data obtained from 258 participants (male and female), where each person has 8 sound files, namely do, re, mi, fa, sol, la, si, and do'. saved in .wav format. From the research conducted, the test was carried out using the K-NN and K-NN methods with Hyperparameters. The results obtained in the form of accuracy of 74% and 81%, so that the Harmonic Product Spectrum (HPS) and K-Nearest Neighbor (K-NN) algorithms give good results for determining the type of human voice.

Corpus linguistic and experimental studies on the meaning-preserving hypothesis in Indonesian voice alternations

Abstract One essential feature of voice alternation is that active and passive clauses centred around a given verb express the same meaning: the “meaning-preserving” hypothesis. One effect of the alternation is the different linking of grammatical relations and semantic roles, which affects the identity of the subject. This paper investigates the meaning-preserving hypothesis in voice alternation in Indonesian from a quantitative usage-based perspective by combining corpus-based data with sentence-production experiment data. It analysed Indonesian caused forward/backward motion verbs and the distribution of their (non-)metaphoric senses in active and passive. The findings demonstrate the frequency effects and sense-sensitivity of voice alternation, such that a given voice type of a verb is strongly associated with certain senses. This finding provides initial support for a previous study on voice alternation in an Austronesian language, predicting that the verb’s semantic properties may condition the statistical bias of the verb towards a particular voice. Some convergence between experimental and corpus findings indicates that participants demonstrate some representation of the strong association between a given voice form of the verb and the sense predominantly expressed in that form, highlighting the notion of item-specific representations of linguistic knowledge as found in construction grammar.

Chaos Behavior Analysis of Alaryngeal Voices Including Esophageal and Tracheoesophageal Voices

Hypothesis/Objectives: This study’s objective was to develop a method to evaluate the chaotic characteristic of alaryngeal speech. The proposed method will be capable of distinguishing between normal and alaryngeal voices, including esophageal (SE) and tracheoesophageal (TE) voices. It has been previously shown that alaryngeal voices exhibit chaotic characteristics due to the aperiodicity of their signals. The proposed method will be applied for future use to quantify both chaos behavior (CB) and the difference between SE and TE voices. Study Design: A total of 74 voice recordings including 34 normal and 40 alaryngeal (26 SE and 14 TE) were used in the study. Voice samples were analyzed to distinguish alaryngeal voices from normal voices and to investigate different chaotic characteristics of SE and TE speech. Methods: A chaotic distribution detection-based method was used to investigate the CB of alaryngeal voices. This CB was used to detect the difference between SE and TE voice types. Quantification of the CB parameter was performed. Statistical analyses were used to compare the results of the CB analysis for both the SE and TE voices. Results: Statistical analysis revealed that CB effectively differentiated between all normal and alaryngeal voice types (p < 0.01). Subsequent multiclass receiver operating characteristic (ROC) analysis demonstrated that CB (area under the curve) possessed the greatest classification accuracy relative to correlation dimension (D₂). Conclusions: The CB metric shows strong promise as an accurate, useful metric for objective differentiation between all normal and alaryngaeal, SE and TE voice types. The CB calculations showed expected results, as SE voices have significantly more CB than TE voices, constituting substantial improvement over previous methods and becoming the first SE and TE classification method. This metric can help clinicians obtain additional acoustic information when monitoring the efficacy of treatment for patients undergoing total laryngectomies.

Design of Intelligent Classification Trash Can System

Intelligent garbage can can identify the name of garbage by voice to identify the type of garbage and accurately classify the type of garbage. This design uses a control system for multi-functional intelligent garbage bins based on STM32F103C8T6. The system mainly consists of speech recognition module, motor drive module, buzzer module, voice broadcast module, power module, infrared obstacle avoidance module and display module, then steering module and embedded system as the core, by reading the garbage type of voice, using language recognition module, through the language broadcast module broadcast garbage type, and then automatically open the corresponding trash can lid, when the trash can reaches a certain height, will make the corresponding prompt. This system can be simple operation, low cost, simple structure, and has the advantages of intelligence, in order to prevent the waste of resources, to prevent waste pollution to the environment, the intelligent voice recognition and classification can be used to solve this problem.

Linguistics Features of Reading Passage in English Text Book K-13 Revised Edition for Senior High School Students in Indonesia: Analysis of Basic Text Properties

The objectives of this research are to identify the linguistic features contained in K 13 English textbook revised edition of grades X, XI, and XII published by the Ministry of Education and Culture of the Republic of Indonesia in terms of sentence pattern, types of tenses, voice, and aspect. This research is descriptive quantitative research. The sample of this research is thirty-nine reading passages were extracted from K13 English textbook revised edition published by the Indonesian Ministry of Education and Culture for senior high school students in Indonesia. All reading passages are taken as a sample, including fourteen texts of grade X, nine texts of grade XI, and sixteen texts of grade XII. The data were gathered by using a checklist and were analyzed by content analysis method. The findings showed that 1) The most common sentence patterns found are simple sentences. 2) Simple past tense is the majority of tenses used in reading the passages, especially at grade X and XI textbooks, while simple present tense is the dominant tenses found at grade XII. 3) The use of active voice dominates all reading passages at all grade levels. 4) The intensity of using the perfective aspect is higher than the imperfective aspect. 5) This result also implies that there are some differences in the findings of linguistic features from the three levels of this book. First, in terms of sentence patterns used in textbooks, at grade X the order of sentence patterns begins from the simple to the complicated one; simple sentences, then compound sentences, next complex sentences, and finally compound-complex sentences. Meanwhile, at grades XI and XII, the most complicated sentence patterns, namely compound-complex sentences, are less than compound sentences. Second, in terms of the types of tenses, the difference is that at grade X and XI textbooks, it is found that the dominant tenses used are simple past tense, while at grade XII textbooks, the most common tenses are simple present tense. these results indicate that the four linguistic features contained in English textbooks for high school In conclusion students in Indonesia need to be taught to students so that students can more easily understand and capture the meanings contained in the textbooks.

Significant pathological voice discrimination by computing posterior distribution of balanced accuracy

The ability to speak lucidly plays a key role in social relations. Consequently, the role of the larynx is quite important, and timely diagnosis of laryngeal diseases has proved to be crucial. In this study, a simple computational model for inverse of speech production model is employed to extract the glottal waveform using speech signal. This waveform has useful information about vocal folds performance in terms of providing evidence for distinguishing pathological disorders. Furthermore, obtaining the significance of classification results is important, because it leads to reliable inferences. This study utilizes the sustained vowel sound /a/ and a well-referenced database, namely MEEI. In this work, after extraction of six discriminating features by using appropriate signal modeling and processing methods and upon change of the feature space by using Kernel Principal Component Analysis (KPCA), a classifier consisting of Naïve Bayes and Fisher Linear Discriminant, is exerted on the feature sets. Regarding voice pathology detection, the proposed approach achieved a significant classification balanced accuracy 93.6%±0.03 with p-value <0.01 for normal and abnormal classification using the Beta distribution model for the posterior distribution of the average of the cross-validation results. The proposed features are also compared with some conventional features in this field. The results show significant improved performance for the proposed features in discriminating different types of pathological voices.

Impact of Vocal Effort on Respiratory and Articulatory Kinematics

Purpose: The goal of this study was to examine the effects of increases in vocal effort, without changing speech intensity, on respiratory and articulatory kinematics in young adults with typical voices. Method: A total of 10 participants completed a reading task under three speaking conditions: baseline, mild vocal effort, and maximum vocal effort. Respiratory inductance plethysmography bands around the chest and abdomen were used to estimate lung volumes during speech, and sensor coils for electromagnetic articulography were used to transduce articulatory movements, resulting in the following outcome measures: lung volume at speech initiation (LVSI) and at speech termination (LVST), articulatory kinematic vowel space (AKVS) of two points on the tongue dorsum (body and blade), and lip aperture. Results: With increases in vocal effort, and no statistical changes in speech intensity, speakers showed: (a) no statistically significant differences in LVST, (b) statistically significant increases in LVSI, (c) no statistically significant differences in AKVS measures, and (d) statistically significant reductions in lip aperture. Conclusions: Speakers with typical voices exhibited larger lung volumes at speech initiation during increases in vocal effort, paired with reduced lip displacements. To our knowledge, this is the first study to demonstrate evidence that articulatory kinematics are impacted by modulations in vocal effort. However, the mechanisms underlying vocal effort may differ between speakers with and without voice disorders. Thus, future work should examine the relationship between articulatory kinematics, respiratory kinematics, and laryngeal-level changes during vocal effort in speakers with and without voice disorders. Supplemental Material: https://doi.org/10.23641/asha.17065457

Describing the voice of online bullying: An analysis of stance and voice type in YouTube comments

In this article, I analyze the expression of stance in YouTube comments responding to a speech by Greta Thunberg addressing the 2019 United Nations Climate Action Summit. I use quantitative and qualitative analysis of explicit stance features (including attitude markers, boosters, hedges, self-mentions, and reader addresses) in order to characterize the voice type that commenters construe and examine how this voice type potentially functions as a tool for social influence. The analysis shows that the comments include continual emphatic and emotive attitude marking, almost exclusively evaluating Thunberg herself (e.g., her authenticity, cognitive ability, gender and youth), rather than directly evaluating her ideas about climate change. Commenters’ evaluations drew on common vocabulary, abbreviations, symbols, and in-group references, which were boosted for example through repetition, capitalization, expletives, and superlatives. As the evaluations were almost uniformly uncivil, positioning Thunberg as a figure of ridicule and hate, I argue that they constitute a recognizable bullying voice that has become common on social media polylogues, especially those concerned with news or politics. This voice type seems to function as a means of constructing a sense of collective identity and purpose, and as an overt show of group power. The ridicule is co-constructed by commenters as a mass social judgement, with the ‘we’ of the commenters consisting of ‘everyone’, albeit with terms of address more typically used for boys/men.

Impersonals, passives, and impersonal pronouns: Lessons from Lithuanian

AbstractThis study examines the properties of VoiceP and impersonal pronouns by contrasting two constructions in Lithuanian: the‐ma/‐taimpersonal and the canonical passive. I argue that, while these two constructions overlap morphologically, they are syntactically distinct. The‐ma/‐taimpersonal is related to the‐no/‐toconstruction in Polish and in Ukrainian. Although it patterns with the Ukrainian‐no/‐topassive in allowing an auxiliary, it behaves like an active VoiceP with a null projected initiator in a thematic‐subject position, a pattern found in the Polish‐no/‐toimpersonal and other impersonals crosslinguistically. I show that the Lithuanian passive lacks a syntactically realized initiator and selects for a type of Voice without a specifier. I also analyze the properties of the impersonal pronoun of the‐ma/‐taimpersonal, demonstrating that it is a bare N that lacks inherently specified ɸ features (number, gender, and person) and has no case. This finding supports existing proposals that treat impersonal pronouns in different languages as defective.

MELODY AS A FACTOR OF VOCAL AND PERFORMING INTERPRETATION OF THE OPERA IMAGE

The aim of the article is to determine current approaches to the phenomenon of opera melodics, clarify the concept of opera melody, substantiate the priority of vocal and performing interpretation in the formation of an opera image and in the development of an opera musical text. A comprehensive methodology for studying the genre nature of opera and its typological artistic features is proposed. The combination of textual and semantic approaches makes it possible to expand the understanding of the melodic functions of an opera text and its communicative tasks. The scientific novelty of the research is due to the identification of aesthetic, structural and compositional, artistic and psychological prerequisites for the creation of the theory of opera melos as an autonomous discipline that can influence modern opera studies and the creative practice of opera performers. The conclusions of the article indicate the key properties of the opera melody, which make it possible to characterize it as a way of creating an opera speech; emphasize the main factors of the opera melodic speech: the enlargement of the elements of the theme and the strengthening of its discreteness, the altitude-rhythmic fixation of the motivational relief, which makes it possible to convey the sound and content of the word, orientation on the “image of feeling”, emotional fullness and psychological personal reasoning, the requirement of getting used to and rebirth into an opera image and specific timbre predestination, that is, being attached to a certain type of vocal voice.

Relationships Among Personality, Daily Speaking Voice Use, and Phonotrauma in Adult Female Singers.

Purpose:This study sought to determine whether personality traits related to extraversion and impulsivity are more strongly associated with singers with nodules compared to vocally healthy singers and to understand the relationship between personality and the types of daily speaking voice use.Method:Weeklong ambulatory voice recordings and personality inventories were obtained for 47 female singers with nodules and 47 vocally healthy female singers. Paired t tests investigated trait differences between groups. Relationships between traits and weeklong speaking voice measures (vocal dose, sound pressure level [SPL], neck surface acceleration magnitude [NSAM], fundamental frequency, cepstral peak prominence [CPP], and the ratio of the first two harmonic magnitudes [H 1 –H 2]) were examined using pairwise Pearson r coefficients. Multiple regressions were performed to estimate voice parameters that correlated with two or more traits.Results:Singers with nodules scored higher on the Social Potency scale (reflecting a tendency toward social dominance) and lower on the Control scale (reflecting impulsivity) compared to the vocally healthy singers. In vocally healthy singers, vocal dose measures were positively correlated with a combination of Wellbeing (i.e., happiness) and Social Potency, mean SPL was positively correlated with Wellbeing, SPL variability was positively correlated with Social Potency and negatively with Harm Avoidance, and CPP mean was positively correlated with Wellbeing. Singers with nodules had a negative correlation between NSAM skewness and Social Potency. Both groups had negative correlations between H 1 –H 2 mean and Social Potency and Social Closeness.Conclusions:Singers with nodules are more socially dominant and impulsive than vocally healthy singers. Personality traits are related to daily speaking voice use, particularly in vocally healthy singers. Individuals with higher levels of traits related to happiness and social dominance and lower Harm Avoidance tended to speak more, with higher laryngeal forces, with more SPL variability, and with more pressed glottal closure, which could increase risk of phonotrauma.

A Study of Zero Anaphora Resolution in Chinese Discourse: From the Perspective of Psycholinguistics.

Anaphora is the main linguistic means to establish discourse coherence, and anaphora resolution is the psychological process to maintain this coherence. Chinese discourse is characterized with providing multiplicity of linguistic clues to readers by employing various referential apparatuses such as pronoun anaphora, zero anaphora, and so on. As a way of avoiding repeated reference to an object that is mentioned beforehand, zero anaphora is frequently employed in discourse. The production and resolution of zero anaphora largely concerns some constraints underlying psychological mechanisms. We particularly focus on zero anaphora resolution in the present study to try to discover some specific aspects of the underlying mechanism, hoping to find out some factors unique to the resolution process. We designed the first two experiments to probe into the psychological reality when participants were presented with sentences containing either pronoun anaphora or zero anaphora or both under discourse condition with topic continuity in Experiment 1a and topic discontinuity in Experiment 1b. We did not find any significant difference in the reaction time between zero anaphora resolution and pronoun anaphora resolution, indicating that zero anaphora possibly works within the processing mechanism on which pronoun anaphora resolution depends. However, we found significantly longer time in reading the first sentence in any of the discourse, showing that the first-mention effect exists in anaphora resolution. We further explored the time course of zero anaphora resolution by measuring the reaction time during the period when participants read sentences that varied according to the location where zero anaphora occurred under two conditions: topic continuity (Experiment 2a) vs. topic discontinuity (Experiment 2b). The strategies of searching for the referential information were found divergent: the exhaustive searching strategy was adopted when the topics within a discourse were kept continuous and the heuristic searching strategy was employed when the topics were discontinuous. The design of Experiment 5 took the factor of voice type and situation consistency into consideration, investigating in what way do those factors influence the resolution of zero anaphora. The voice type, according to the results, plays a significant role for its exclusively close relationship with the first-mention effect.