The effects of talker familiarity on spoken word identification in younger and older listeners

The present study was designed to examine age differences in the ability to use voice information acquired intentionally (Experiment 1) or incidentally (Experiment 2) as an aid to spoken word identification. Following both implicit and explicit voice learning, participants were asked to identify novel words spoken either by familiar talkers (ones they had been exposed to in the training phase) or by 4 unfamiliar voices. In both experiments, explicit memory for talkers' voices was significantly lower in older than in young listeners. Despite this age-related decline in voice recognition, however, older adults exhibited equivalent, and in some cases greater, benefit than young listeners from having words spoken by familiar talkers. Implications of the findings for age-related changes in explicit versus implicit memory systems are discussed.     

The processing of auditory and visual recognition of self-stimuli

This study examined self-recognition processing in both the auditory and visual modalities by determining how comparable hearing a recording of one’s own voice was to seeing photograph of one’s own face. We also investigated whether the simultaneous presentation of auditory and visual self-stimuli would either facilitate or inhibit self-identification. Ninety-one participants completed reaction-time tasks of self-recognition when presented with their own faces, own voices, and combinations of the two. Reaction time and errors made when responding with both the right and left hand were recorded to determine if there were lateralization effects on these tasks. Our findings showed that visual self-recognition for facial photographs appears to be superior to auditory self-recognition for voice recordings. Furthermore, a combined presentation of one’s own face and voice appeared to inhibit rather than facilitate self-recognition and there was a left-hand advantage for reaction time on the combined-presentation tasks.     

Recalling semantic information about newly learned faces and voices

Several findings showed that semantic information is more likely to be retrieved from recognised faces than from recognised voices. Earlier experiments, which investigated the recall of biographical information following person recognition, used stimuli that were pre-experimentally familiar to the participants, such as famous people's voices and faces. We propose an alternative method to compare the participants’ ability to associate semantic information with faces and voices. The present experiments allowed a very strict control of frequency of exposure to pre-experimentally unfamiliar faces and voices and ensured the absence of identity clues in the spoken extracts. In Experiment 1 semantic information was retrieved from the presentation of a name. In Experiment 2 semantic and lexical information was retrieved from faces and/or voices. A memory advantage for faces over voices was again observed.     

Commonsense beliefs and the identification of familiar voices

Two experiments are reported in which participants attempted to reject the tape‐recorded voice of a stranger and identify by name the voices of three personal associates who differed in their level of familiarity. In Experiment 1 listeners were asked to identify speakers as soon as possible, but were not allowed to change their responses once made. In Experiment 2 listeners were permitted to change their responses over successive presentations of increasing durations of voice segments. Also, in Experiment 2 half of the listeners attempted to identify speakers who spoke in normal‐tone voices, and the remainder attempted to identify the same speakers who spoke in whispers. Separate groups of undergraduate students attempted to predict the performance of the listeners in both experiments. Accuracy of performance depended on the familiarity of speakers and tone of speech. A between‐subjects analysis of rated confidence was diagnostic of accuracy for high familiar and low familiar speakers (Experiment 1), and for moderate familiar and unfamiliar normal‐tone speakers (Experiment 2). A modified between‐subjects analysis assessed across the four levels of familiarity yielded reliable accuracy‐confidence correlations in both experiments. Beliefs about the accuracy of voice identification were inflated relative to the significantly lower actual performance for most of the normal‐tone and whispered‐speech conditions. Forensic significance and generalizations are addressed. Copyright © 2001 John Wiley & Sons, Ltd.     

Perceptual prioritization of self-associated voices

Information associated with the self is prioritized relative to information associated with others and is therefore processed more quickly and accurately. Across three experiments, we examined whether a new externally-generated voice could become associated with the self and thus be prioritized in perception. In the first experiment, participants learned associations between three unfamiliar voices and three identities (self, friend, stranger). Participants then made speeded judgements of whether voice-identity pairs were correctly matched, or not. A clear self-prioritization effect was found, with participants showing quicker and more accurate responses to the newly self-associated voice relative to either the friend- or stranger- voice. In two further experiments, we tested whether this prioritization effect increased if the self-voice was gender-matched to the identity of the participant (Experiment 2) or if the self-voice was chosen by the participant (Experiment 3). Gender-matching did not significantly influence prioritization; the self-voice was similarly prioritized when it matched the gender identity of the listener as when it did not. However, we observed that choosing the self-voice did interact with prioritization (Experiment 3); the self-voice became more prominent, via lesser prioritization of the other identities, when the self-voice was chosen relative to when it was not. Our findings have implications for the design and selection of individuated synthetic voices used for assistive communication devices, suggesting that agency in choosing a new vocal identity may modulate the distinctiveness of that voice relative to others.     

Auditory Long term Memory: Repetition Priming of Voice Recognition

Two experiments examined repetition priming in the recognition of famous voices. In Experiment 1, reaction times for fame decisions to famous voice samples were shorter than in an unprimed condition, when voices were primed by a different voice sample of the same person having been presented in an earlier phase of the experiment. No effect of voice repetition was observed for non-famous voices. In Experiment 2, it was investigated whether this priming effect is voice-specific or whether it is related to post-perceptual processes in person recognition. Recognizing a famous voice was again primed by having earlier heard a different voice sample of that person. Although an earlier exposure to that person's name did not cause any priming, there was some indication of priming following an earlier exposure to that person's face. Finally, earlier exposure to the identical voice sample (as compared to a different voice sample from the same person) caused a considerable bias towards responding 'famous'-i.e. performance benefits for famous but costs for nonfamous voices. The findings suggestthat (1) repetition priming invoice recognition primarily involves the activation of perceptual representations of voices, and (2) it is important to determine the conditions in which priming causes bias effects that need to be disentangled from performance benefits.     

Why are familiar-only experiences more frequent for voices than for faces?

Hanley,Smith, and Hadfield (1998) showed that when participants were asked to recognize famous people from hearing their voice , there was a relatively large number of trials in which the celebrity's voice was felt to be familiar but biographical information about the person could not be retrieved. When a face was found familiar, however, the celebrity's occupation was significantly more likely to be recalled. This finding is consistent with the view that it is much more difficult to associate biographical information with voices than with faces. Nevertheless, recognition level was much lower for voices than for faces in Hanleyet al.'s study,and participants made significantly more false alarms in the voice condition. In the present study, recognition performance in the face condition was brought down to the same level as recognition in the voice condition by presenting the faces out of focus. Under these circumstances, it proved just as difficult to recall the occupations of faces found familiar as it was to recall the occupations of voices found familiar. In other words, there was an equally large number of familiar-only responses when faces were presented out of focus as in the voice condition. It is argued that these results provide no support for the view that it is relatively difficult to associate biographical information with a person's voice. It is suggested instead that associative connections between processing units at different levels in the voice-processing system are much weaker than is the case with the corresponding units in the face-processing system. This will reduce the recall of occupations from voices even when the voice has been found familiar. A simulation was performed using the latest version of the IAC model of person recognition (Burton, Bruce, & Hancock, 1999) which demonstrated that the model can readily accommodate the pattern of results obtained in this study.     

Crossmodal binding: evaluating the "unity assumption" using audiovisual speech stimuli

We investigated whether the "unity assumption," according to which an observer assumes that two different sensory signals refer to the same underlying multisensory event, influences the multisensory integration of audiovisual speech stimuli. Syllables (Experiments 1, 3, and 4) or words (Experiment 2) were presented to participants at a range of different stimulus onset asynchronies using the method of constant stimuli. Participants made unspeeded temporal order judgments regarding which stream (either auditory or visual) had been presented first. The auditory and visual speech stimuli in Experiments 1-3 were either gender matched (i.e., a female face presented together with a female voice) or else gender mismatched (i.e., a female face presented together with a male voice). In Experiment 4, different utterances from the same female speaker were used to generate the matched and mismatched speech video clips. Measuring in terms of the just noticeable difference the participants in all four experiments found it easier to judge which sensory modality had been presented first when evaluating mismatched stimuli than when evaluating the matched-speech stimuli. These results therefore provide the first empirical support for the "unity assumption" in the domain of the multisensory temporal integration of audiovisual speech stimuli.     

Recalling semantic information about newly learned faces and voices

Several findings showed that semantic information is more likely to be retrieved from recognised faces than from recognised voices. Earlier experiments, which investigated the recall of biographical information following person recognition, used stimuli that were pre-experimentally familiar to the participants, such as famous people's voices and faces. We propose an alternative method to compare the participants' ability to associate semantic information with faces and voices. The present experiments allowed a very strict control of frequency of exposure to pre-experimentally unfamiliar faces and voices and ensured the absence of identity clues in the spoken extracts. In Experiment 1 semantic information was retrieved from the presentation of a name. In Experiment 2 semantic and lexical information was retrieved from faces and/or voices. A memory advantage for faces over voices was again observed.     

Recalling semantic information about personally known faces and voices

Previous research that investigated whether biographical information about familiar people is harder to retrieve from voices than from faces produced contrasting results. However, studies that used a strict control of the content of spoken extracts reported that semantic information about familiar people is easier to retrieve when recognising a face than when recognising a voice. In all previous studies faces and voices of famous people were used as stimuli. In the present study, personally familiar people's voices and faces (standard faces and blurred faces) were used. Presenting such people (i.e., participants’ teachers) allowed controlling still more strictly the content of the spoken extracts since it was possible to ask all the target persons to speak the same words. In addition, it was previously stressed that we encounter famous people's faces in the media more frequently than we hear their voice. This methodological difficulty was presumably reduced when teachers’ faces were presented. Present results showed a significant decrease in retrieval of biographical information from familiar voices relative to blurred faces even though the level of overall recognition was similar for blurred faces and voices. The role of the relative distinctiveness of voices and faces is discussed and further investigation is proposed.     

Reading voices and hearing text: talker-specific auditory imagery in reading

A series of experiments was conducted to determine if linguistic representations accessed during reading include auditory imagery for characteristics of a talker's voice. In 3 experiments, participants were familiarized with two talkers during a brief prerecorded conversation. One talker spoke at a fast speaking rate, and one spoke at a slow speaking rate. Each talker was identified by name. At test, participants were asked to either read aloud (Experiment 1) or silently (Experiments 1, 2, and 3) a passage that they were told was written by either the fast or the slow talker. Reading times, both silent and aloud, were significantly slower when participants thought they were reading a passage written by the slow talker than when reading a passage written by the fast talker. Reading times differed as a function of passage author more for difficult than for easy texts, and individual differences in general auditory imagery ability were related to reading times. These results suggest that readers engage in a type of auditory imagery while reading that preserves the perceptual details of an author's voice.     

The role of audiovisual asynchrony in person recognition

Audiovisual integration (AVI) has been demonstrated to play a major role in speech comprehension. Previous research suggests that AVI in speech comprehension tolerates a temporal window of audiovisual asynchrony. However, few studies have employed audiovisual presentation to investigate AVI in person recognition. Here, participants completed an audiovisual voice familiarity task in which the synchrony of the auditory and visual stimuli was manipulated, and in which visual speaker identity could be corresponding or noncorresponding to the voice. Recognition of personally familiar voices systematically improved when corresponding visual speakers were presented near synchrony or with slight auditory lag. Moreover, when faces of different familiarity were presented with a voice, recognition accuracy suffered at near synchrony to slight auditory lag only. These results provide the first evidence for a temporal window for AVI in person recognition between approximately 100 ms auditory lead and 300 ms auditory lag.     

Limitations of episodic memory for highly similar auditory stimuli

Two experiments investigated participants’ recognition memory for word content, while varying vocal characteristics, and for vocal characteristics alone. In Experiment 1, participants performed an auditory recognition task in which they identified whether a spoken word was “new”, “old” (repeated word, repeated voice), or “similar” (repeated word, new voice). Results showed that word recognition accuracy was lower for similar trials than old trials. In Experiment 2, participants performed an auditory recognition task in which they identified whether or not a phrase was spoken in an old or new voice, with repetitions occurring after a variable number of intervening stimuli. Results showed that recognition accuracy was lower when old voices spoke an alternate message than a repeated message and accuracy decreased as a function of number of intervening items. Overall, the results suggest that speech recognition is better for lexical content than vocal characteristics alone.     

Crossmodal binding: Evaluating the “unity assumption” using audiovisual speech stimuli

We investigated whether the “unity assumption,” according to which an observer assumes that two different sensory signals refer to the same underlying multisensory event, influences the multisensory integration of audiovisual speech stimuli. Syllables (Experiments 1, 3, and 4) or words (Experiment 2) were presented to participants at a range of different stimulus onset asynchronies using the method of constant stimuli. Participants made unspeeded temporal order judgments regarding which stream (either auditory or visual) had been presented first. The auditory and visual speech stimuli in Experiments 1–3 were either gender matched (i.e., a female face presented together with a female voice) or else gender mismatched (i.e., a female face presented together with a male voice). In Experiment 4, different utterances from the same female speaker were used to generate the matched and mismatched speech video clips. Measuring in terms of the just noticeable difference the participants in all four experiments found it easier to judge which sensory modality had been presented first when evaluating mismatched stimuli than when evaluating the matched-speech stimuli. These results therefore provide the first empirical support for the “unity assumption” in the domain of the multisensory temporal integration of audiovisual speech stimuli.     

Objective self awareness and self esteem

Based on a theory of objective self awareness three experiments were conducted with undergraduates to test the hypothesis that self-focused attention can alter self esteem levels. In Experiments I and II subjects were exposed either to the sound of their own voices or to the sound of another's voice, and while listening to the tape-recording they filled out a self esteem measure. Consistent with the theory, subjects whose attention was focused upon themselves by means of exposure to their own tape-recorded voices showed lower self esteem than subjects who heard another's voice. The impact of the self awareness manipulation on self esteem was greatest when it was first introduced. In Experiment III the variable of positive-negative feedback on a fictitious personality trait was added to the self-focused attention variable. An interaction resulted such that self-focused attention lowered self esteem given negative feedback, while there was a tendency toward the opposite result given positive feedback.     

The effect of familiarity on within-person age judgements from voices

Listeners can perceive a person’s age from their voice with above chance accuracy. Studies have usually established this by asking listeners to directly estimate the age of unfamiliar voices. The recordings used mostly include cross-sectional samples of voices, including people of different ages to cover the age range of interest. Such cross-sectional samples likely include not only cues to age in the sound of the voice but also socio-phonetic cues, encoded in how a person speaks. How age perpcetion accuracy is affected when minimizing socio-phonetic cues by sampling the same voice at different time points remains largely unknown. Similarly, with the voices in age perception studies being usually unfamiliar to listeners, it is unclear how familiarity with a voice affects age perception. We asked listeners who were either familiar or unfamiliar with a set of four voices to complete an age discrimination task: listeners heard two recordings of the same person’s voice, recorded 15 years apart, and were asked to indicate in which recording the person was younger. Accuracy for both familiar and unfamiliar listeners was above chance. While familiarity advantages were apparent, accuracy was not particularly high: familiar and unfamiliar listeners were correct for 68.2% and 62.7% of trials, respectively (chance = 50%). Familiarity furthermore interacted with the voices included. Overall, our findings indicate that age perception from voices is not a trivial task at all times – even when listeners are familiar with a voice. We discuss our findings in the light of how reliable voice may be as a signal for age.     

The role of language familiarity in voice identification

Four experiments examined the effects of language characteristics on voice identification. In Experiment 1, monolingual English listeners identified bilinguals' voices much better when they spoke English than when they spoke German. The opposite outcome was found in Experiment 2, in which the listeners were monolingual in German. In Experiment 3, monolingual English listeners also showed better voice identification when bilinguals spoke a familiar language (English) than when they spoke an unfamiliar one (Spanish). However, English-Spanish bilinguals hearing the same voices showed a different pattern, with the English-Spanish difference being statistically eliminated. Finally, Experiment 4 demonstrated that, for English-dominant listeners, voice recognition deteriorates systematically as the passage being spoken is made less similar to English by rearranging words, rearranging syllables, and reversing normal text. Taken together, the four experiments confirm that language familiarity plays an important role in voice identification.     

A "voice inversion effect?"

Voice is the carrier of speech but is also an "auditory face" rich in information on the speaker's identity and affective state. Three experiments explored the possibility of a "voice inversion effect," by analogy to the classical "face inversion effect," which could support the hypothesis of a voice-specific module. Experiment 1 consisted of a gender identification task on two syllables pronounced by 90 speakers (boys, girls, men, and women). Experiment 2 consisted of a speaker discrimination task on pairs of syllables (8 men and 8 women). Experiment 3 consisted of an instrument discrimination task on pairs of melodies (8 string and 8 wind instruments). In all three experiments, stimuli were presented in 4 conditions: (1) no inversion; (2) temporal inversion (e.g., backwards speech); (3) frequency inversion centered around 4000 Hz; and (4) around 2500 Hz. Results indicated a significant decrease in performance caused by sound inversion, with a much stronger effect for frequency than for temporal inversion. Interestingly, although frequency inversion markedly affected timbre for both voices and instruments, subjects' performance was still above chance. However, performance at instrument discrimination was much higher than for voices, preventing comparison of inversion effects for voices vs. non-vocal stimuli. Additional experiments will be necessary to conclude on the existence of a possible "voice inversion effect."     

Evaluating the effectiveness of pixelation and blurring on masking the identity of familiar faces

Two experiments are reported that assess how well the identity of highly familiar (famous) faces can be masked from short naturalistic television clips. Recognition of identity was made more difficult by either pixelating (Experiment 1) or blurring (Experiment 2) the viewed face. Participants were asked to identify faces from both moving and static clips. Results indicated that participants were still able to recognize some of the viewed faces, despite these image degradations. In addition, moving images of faces were recognized better than static ones. The practical and theoretical implications of these findings are discussed. Copyright © 2001 John Wiley & Sons, Ltd.