Music and adults with CI

Music is both an art form and a social activity making its two primary functions aesthetic enjoyment and entertainment. Currently, research is accumulating regarding music perception and appreciation of implant recipients. A greater understanding of this subject may address such issues as user satisfaction with daily functioning (Stordahl, 2002) and music listening as part of it (Gfeller, Witt, Mehr, Woodworth, &

Knutson, 2002c).

While the processing and transmission of selective aspects of the sound wave have proven quite successful in speech perception by implant recipients, there are serious limitations with regard to the representation of musical sounds. Specifically, cochlear implant recipients do not receive a normal representation of pitch and timbre, which are both important structural features in music (Gfeller et al., 2002a).

For example, there is not an orderly relationship between the frequency (Hz) of the sound source and the pitch perception by the implant recipient of how high or low a note sounds. A person with NH would hear an ascending C major scale as an orderly progression of pitches from low to high, while an implant recipient may hear this same scale as a random array of beeps or tones (Fujita & Ito, 1999; Gfeller, 1996).

This is true for harmony as well (Gfeller, 2001).

Rhythm²⁰, which embraces the temporal features of music, is the musical parameter that is the most readily perceived by CI recipients. The beat unit, tempo, rhythmical patterns, pauses, and meter- all have a cardinal role in music recognition for CI recipients. In a study by Gfeller et al (2002a), the authors compared song recognition and pitch perception of adult CI recipients and adults with NH and tried to identify factors that influence the ability of implant users to recognize familiar

20 The reader may refer to the researcher’s master’s dissertation for detailed explanations of the musical parameters (see section 1.3).

melodies. Forty nine experienced CI recipients and 18 adults with NH were tested on familiar melody recognition. The adults with NH were significantly more accurate than the CI recipients. Two thirds of the melodies that were correctly recognized included a rhythmic line that was highly memorable and served as a cue for the listeners. The CI recipients performed similarly (almost normally) to the adults with NH in tempo discrimination tasks, but showed considerable variability in perception of complex tones and pure tones. In a related study done by Kong, Cruz, Jones, and Zeng (2004), CI recipients performed normally in tempo discrimination, and were able to identify almost two thirds of melodies of familiar songs that contained rhythmic information. However, they were significantly poorer in rhythmic pattern identification and melody recognition compared to the listeners with NH. Melody recognition requires perception of temporal (rhythmic) and spectral (pitch) cues. The CI recipients mostly relied on the rhythmic cues in melody recognition tasks. Fujita and Ito (1999) found that the eight adults with CI recognized the melodies of four songs according to verbal cues rather than by musical qualities such as tones and musical intervals (The closest discrimination was 4 semitones). Leal et al. (2003) asked 29 adults with CI to listen to 10 pairs of musical pieces and discriminate whether the pair of items were the “same” or ”different” as well as determining the point of change. In the 10 pairs presented to them, one was different in the duration or intensity of the notes. A total of 59% of the CI recipients were able to discriminate and 41%- to identify the point of change. The researchers also found positive correlations between speech discrimination and rhythm tasks.

Pitch perception is crucial for identification and enjoyment of music. Melodies, which are part of music, are created when a sequence of pitches (single tones) are successively organized to build a phrase of varying musical contour and interval.

For a CI recipient to hear and appreciate music in the same way as a person with NH does, he/she must be able to discriminate between subtle changes in pitch (frequencies) as well as musical intervals (Limb, 2000). It appears that many implant users can perform discriminations of pitch and rhythms, especially when electric pulse stimuli are applied (Gfeller, Knutson, Woodworth, Witt, & DeBus, 1998; Pijl, 1997; McDermott & McKay, 1997). However, pitch and timbre information through speech processors is not always perceived with precision (Limb, 2000). Fujita and Ito (1999) indicate that there is a great variability among CI recipients in their ability to

discriminate musical intervals, from discriminating changes of four semitones to inability to identify changes of an octave. (See also the study by Gfeller et al. (2002a), in the previous paragraph, on song recognition and pitch perception).

Timbre, or tone quality or color, is another fundamental feature of music.

Recognition of a musical instrument is not required to derive pleasure from music.

What is more important is the satisfaction of the CI recipients regarding the extent to which the sound is pleasing (Gfeller et al, 2002c). Therefore, Gfeller and other authors examined the appraisal and recognition of timbre (four different musical instruments) by adults with CI compared to adults with NH (Gfeller et al., 1998). The latter were able to recognize all of the instruments with significantly greater accuracy than implant recipients. Performance on timbre perception tasks was correlated with speech and cognitive tasks. Results from another study (Gfeller et al., 2002c) support this study as well. In the study of Gfeller et al., subjects were asked to appraise the overall pleasantness of recorded instruments (a 7-note sequence of equal-duration notes). Whereas prior research indicates that adults with NH tend to rate musical sounds with more high-frequency energy with the verbal descriptor “brilliant”, the CI recipients rated it as significantly more scattered (i.e., noisy) and dull in tone quality than did the adults with NH. The authors relate these findings to the limitations of the implant in transmitting key structural features of music as well as to the poorer nerve survival in the basal region of the cochlea, which is stimulated by the implant in response to higher-frequency sounds.

One more finding on timbre was that adults with CI showed greater ability to identify percussive instruments, such as the piano, in comparison with woodwind or brass instruments (Gfeller et al., 1998; Gfeller et al., 2002c; Gfeller et al., 2002b).

Additional numerous studies investigated the ability to discriminate between different musical instruments, i.e., different timbres (e.g., Fujita & Ito, 1999; Gfeller

& Lansing, 1991; Leal et al., 2003).

Other studies aimed to determine perceptions of rhythm, pitch, melody, and timbre found that-

• CI adult recipients scored better on rhythm as opposed to tonal and harmonic perception tests (subjects with NH showed higher scores for melodic perception).

• They exhibited greater accuracy for music with “structured” rhythms as opposed to unstructured rhythms.

• They preferred isolated rhythmic patterns over rhythmic with harmonic patterns and were less accurate in a same- different comparison when melodic patterns included small pitch changes (Gfeller & Lansing, 1991;

Gfeller, Woodworth, Robin, Witt, & Knutson, 1997).

McDermott (2004) summarizes the most significant findings of past research on music perception:

1. Persons with CI, on average, perceive rhythm on a par with persons with NH.

2. Although technological advancements of sound processors have been made in the last decade, this has had little impact on improving CI users’ recognition of melodies that are not accompanied by rhythmic or verbal cues.

3. Determining perception of timbre through procedures that require CI users to identify sounds of musical instruments has not proved satisfactory.

4. The quality of musical sounds is rated as less pleasant by people with CI than by those with NH.

5. Through use of auditory training programs created specifically for presentation of structural musical listening experiences to people with CI, an improvement may come about in the subjective acceptability of music that is heard via the CI.

6. Innovative sound processors using both temporal and spatial patterns of electrical stimulation may improve pitch perception by overcoming current limitations of signal coding in the existing implant systems.

7. CI recipients with adequate residual hearing in the low frequency range may receive a combination of electrical and acoustic stimulation. This is likely to improve their perception of music.

Rubinstein (2007)²¹ notes, that there is currently no evidence that CI devices manufactured by different companies differ in their music perception outcomes. That does not mean that there is no difference, but just that no validated study has

21 Prof. J. T. Rubinstein, University of Washington, Seattle (personal communication, December 18th, 2007). E-mail: rubinj@u.washington.edu

compared devices in a large population. In addition, Leal et al. (2003) point out that no studies today have proved an advantage in music perception for bilateral CI users.

The previous paragraphs related to specific isolated structural features of music, but consideration should be given to the experience of music as a gestalt, which combines rhythm, melody, harmony and timbre.

Music listening and enjoyment is problematic for many adult CI recipients, who describe the sound quality of music as scratchy, squeaky, tinny, booming, unnatural, mechanical, or noise-like (Gfeller, Chrsit, Knutson, Witt, & Mehr., 2003). Similarly, Leal et al. (2003) write that generally speaking, listening to music is unpleasant after cochlear implantation. In their study, 25 CI recipients (86%) presented lower scores of listening habits after implantation, and 11 (38%) stated that they did not like listening to music. The findings are supported by the results of a study by Gfeller et al. (2000), in which 23% of subjects reported little satisfaction listening to music prior to or after implantation; 43% reported that the sound of music improved over time or was better than no music at all (although less pleasant than before they lost their hearing) and 23% noted that music sounded as pleasant as before hearing loss or better. In the study of Leal et al., the CI recipients were presented with musical stimuli that conformed as closely as possible to those that occur in everyday music listening situations. Eisenberg (1982) indicates that postlingually deaf adults are often disappointed when hearing music through the implant as the signal lacks the tonal/harmonic structure which hearing persons expect. Those adults have developed a concept or internal mental schemata of what music sounds like prior to losing their hearing. Their recall, however, can sometimes contribute to a highly critical assessment of sound quality, especially in cases of recent hearing loss, when the person still has a vivid recollection of sound quality. In contrast, persons who are prelingually deaf often found music to be enjoyable through the implant and identified music listening as a major motivating factor behind acceptance of the implant. In a study by Mirza et al., appreciation of music after implantation was varied but on the whole, it was disappointing, as evidenced by a questionnaire completed by 35 post-lingual adults with CI (ages 24-81 years, with a mean age of 53). Most of the subjects (54%) did not listen to music. Those who enjoyed it more after the implantation were the younger ones, who had shorter length of deafness and high speech perception scores (Mirza, Douglas, Linsey, Hildreth, & Hawthorne,

2003). Seven subjects, who played a musical instrument before becoming deaf, subsequently listened to music after implantation. Similarly, in an article by Dorman et al., a 35-year-old woman with CI who became deaf at the age of 31 (Dorman, Basham, McCandless, & Dove, 1991), expressed her high satisfaction regarding the implantation and her ability to appreciate music. She recognized familiar tunes on the radio, and continued to be a piano teacher. These findings suggest that musically trained CI recipients are better able to recognize music since they use their experience to extract maximal information from fragmentary musical evidence (Fujita & Ito, 1999).

Implant recipients rate the quality of solo musical instruments poorer than adults with NH (Gfeller et al., 1998; Gfeller & Lansing, 1991, 1992; Gfeller et al., 1997).

Gfeller et al. (2003) indicate that seldom do people hear isolated elements of music in everyday life because music is experienced as complex patterns combining pitch (melody and harmony), timbre and rhythm. In order to determine the effects of prior song familiarity and subjective complexity²² on complex song appraisal, the researchers developed a test that measures appraisal of complex musical stimuli (classical music, country western and pop) of the sort heard in everyday life. Adults with CI (postlingualy deafened) gave significantly less positive ratings to classical music than did adults with NH. They also rated two of three musical genres (country western and pop) as significantly more complex. The authors conclude that classical music is perceived by the implantees as more complex than the songs because subjectively it is more complex (complex rhythmic structures, more complex harmonic changes, contrapuntal organization with multiple melodic themes, etc.) even when peripheral hearing mechanism functions normally. In contrast to classical music, songs include lyrics that can be extracted by many CI recipients (not all of them) as linguistic information. Thus, it may be easier for them to follow the acoustic information. In addition, these songs had a strong, simple beat, which is effectively transmitted by implants.

Gfeller et al. (2000) indicate that mere experience with the device is not associated with improved music perception or enjoyment. Regarding music enjoyment, the researchers found significant correlations between self-report of

22 Subjective complexity: The “perceived complexity level or information content, which is mutable and a function of the listener and past musical experience” (Price, p. 186, as cited in Gfeller et al., 2003).

musical enjoyment and the amount of time devoted to music listening post-implantation.

In order to find whether different types of CI provide CI recipients with more positive or negative music listening experiences, Gfeller et al. (2000) compared four different types (Clarion, Nucleus, MED-EL and Ineraid) and concluded that none of these devices was identified with particularly positive or negative music listening experiences for any of the factors evaluated.

The present CI provides sufficient spectral cues for speech recognition perception, but it is not adequate to support music perception. This may be achieved in the future by increasing the number of functional channels and improving the encoding of the fine structure information (Kong et al., 2004).