Effect of Sound Familiarity on the Event-Related Potentials Elicited by Novel Environmental Sounds

36, 30–51 (1998) BR970955

BRAIN AND COGNITION ARTICLE NO.

Effect of Sound Familiarity on the Event-Related Potentials Elicited by Novel Environmental Sounds Yael M. Cycowicz and David Friedman Cognitive Electrophysiology Laboratory, New York State Psychiatric Institute, New York City, New York The effect of sound familiarity was examined within the context of an eventrelated potential (ERP) novelty oddball paradigm. Brain electrical activity was recorded while subjects (16 young adults) listened to frequent tones, infrequent target tones, and infrequent novel environmental sounds. Subjects were instructed to press a button in response to the target tones only. There were 48 different novel sounds, 32 of which were repeated, and about two-thirds of which represented familiar sound concepts. The novel sounds elicited two ERP components, the novelty P3 and the P32. The novelty P3 was modulated by both repetition and familiarity, such that repeated familiar sounds elicited decreased novelty P3 amplitude at frontal sites, while repeated unfamiliar sounds elicited increased novelty P3 amplitude at posterior sites. This differential effect may reflect the operation of a neural network that distinguishes among different degrees of novelty.  1998 Academic Press

It is well known that mental processes involving word stimuli are affected by word frequency and/or word familiarity. Familiarity accounts for the relative accessibility of a stimulus’s lexical representation and thus determines processing efficiency. For example, words that are more often used in written language are recognized more quickly than rare words (Forster & Chambers, 1973). It is also well documented in the vast priming literature that word repetition improves performance, such as eliciting faster reaction time and greater accuracy compared to unrepeated words (e.g., Scarborough, Cortese, & Scarborough, 1977). The effects of word frequency and word repetition interact, so that low-frequency words benefit more from repetition than do high-frequency words (Forster & Davis, 1984; Jacoby & Dallas, 1981). The authors thank Mr. Charles L. Brown III for computer programming and Ms. Mairav Rothstein for assistance in data collection. We are grateful to all volunteers who generously gave their time to participate in the study. This research was supported in part by Grant HD14959 to David Friedman and by the New York State Department of Mental Hygiene. Address correspondence and reprint request to Dr. Yael M. Cycowicz, Cognitive Electrophysiology Laboratory, New York State Psychiatric Institute, 722 West 168th Street, Unit 58, New York, NY 10032. Fax: (212) 543-6002. E-mail: [email protected]. 30 0278-2626/98 $25.00 Copyright  1998 by Academic Press All rights of reproduction in any form reserved.

SOUND FAMILIARITY AND NOVELTY ERP

31

Although the research data based on nonlinguistic sounds are much less extensive than those for verbal stimuli, there is some evidence suggesting similarities in the ways in which sounds and words are processed. For example, sound familiarity is one determinant of sound identifiability, in that commonly heard sounds are easier to identify than less common sounds (Ballas, 1993). Similarly, using environmental sounds, Stuart and Jones (1995) demonstrated an analogy to the word priming effect. In their study, identification of auditorially presented environmental sounds was facilitated by prior exposure to the same sound. Furthermore, Van Petten and Rheinfelder (1995), who used meaningful nonspeech sounds and related words, demonstrated semantic priming of both words and sounds regardless of whether the nonspeech sound or the related word served as prime or target. While investigations of the effects of familiarity and/or repetition have involved behavioral measures, physiological activity can provide complementary information about the nature of stimulus processing. One such physiological measure is the event-related brain potential (ERP), which is a noninvasive method of recording voltage changes elicited within the brain from the scalp by means of signal averaging. The ERP is a consequence of the activity of neuronal populations that is time-locked to environmental stimuli. The positive and negative ERP voltage fluctuations are known as components and have been shown to reflect specific stages of information processing (review by Picton & Hillyard, 1988). The P300, originally discovered by Sutton, Braren, Zubin, and John (1965), and Sutton et al. (1967), is one such component whose latency reflects the duration of stimulus evaluation (e.g., Duncan-Johnson & Kopell, 1981; McCarthy & Donchin, 1981; review by Johnson, 1986) and is elicited typically by infrequent task-relevant events. Polich and Donchin (1988) hypothesized that if the word frequency effect stemmed from a difference in encoding time between common and uncommon words, then the P300 elicited by common words should have a shorter latency than the P300 elicited by uncommon words. Thus, Polich and Donchin (1988) designed a lexical decision task in which subjects were shown a random sequence of high- and low-frequency words as well as nonwords and were asked to press one button when they saw a word and another button when they saw a nonword. As was expected, the latency of P300 elicited by the high-frequency words was shorter than the latency of P300 elicited by the low-frequency words, and both were shorter than their associated reaction times. These results confirmed the suggestion that the word frequency effect began during the early stages of word processing before the response production stage, and that lexical access for uncommon words took more time than lexical access for common words (Polich & Donchin, 1988). Several investigators have shown that word repetition modulates ERP activity (e.g., Friedman, 1990; Nagy & Rugg, 1989; Rugg, 1987). For example, Rugg (1990) recorded the N400, a negative deflection that peaks at about 400 ms after stimulus onset, to investigate the modulation of ERP amplitudes

32

CYCOWICZ AND FRIEDMAN

by the repetition of high- and low-frequency words. The amplitude of N400 is affected by the extent to which the attributes of the evoking word are predictable within its semantic context (Kutas & Hillyard, 1984), and thus it has been suggested that the N400 reflects postlexical processing (Halgren & Smith, 1987). Rugg (1990) demonstrated that, as in the behavioral studies briefly reviewed above, word frequency and repetition modulated the N400 component of the ERPs in an interactive fashion. In his study, subjects were presented with nonwords, high-frequency words, and low-frequency words. Half of the high- and low-frequency words were repeated, and subjects were asked to respond only to nonwords. On first presentation, low-frequency words evoked the largest N400, while repetition induced a relatively larger reduction in N400 amplitude for low- than for high-frequency words. Thus, it appeared that N400 amplitude was inversely related to word familiarity. The level of stimulus familiarity can be based either upon the frequency with which the stimulus has been encountered outside the experiment or upon the repetition of the stimulus item within the experiment. Therefore, Rugg suggested that the interactive effects of frequency and repetition on N400 reflected the influence of both kinds of familiarity on postlexical processing. Similar results were reported in another study in which a semantic classification task was employed (Young & Rugg, 1992). In the present study the effect of sound familiarity was examined within an ERP novelty oddball paradigm (a variant of the standard oddball task), a paradigm first introduced by Courchesne, Hillyard, and Galambos (1975). In the standard oddball paradigm, frequent and rare or target (oddball) stimuli are randomly presented and subjects are asked to respond in some fashion (either to count or to press a button) to these target stimuli. In this paradigm, ERPs to the attended, task-relevant, infrequent target stimuli elicit a P3 component, which is synonymous with the well-known P300 (also labeled P3b), that reaches peak latency between 300 and 1000 ms poststimulus, depending upon the complexity of the task. The scalp topography of this P3 is usually, but not always, characterized by a parietally focused amplitude maximum (see for example, Johnson, 1993). By contrast, ERPs to unattended infrequent stimuli elicit a short-latency P3 component (at about 280 ms), labeled the P3a (Squires, Squires, & Hillyard, 1975). The P3a shows an amplitude maximum at the midline central scalp site (i.e., Cz). In the novelty oddball task, the frequent and rare target stimuli are intermixed with equally rare stimuli that are unexpected and novel. As in the regular oddball task, subjects are asked to respond to the target stimuli, but are not informed about the presence of the novel events. The P3 elicited by these novel events has a more frontally oriented scalp topography than the P3b (e.g., Courchesne et al., 1975; Cycowicz, Friedman, & Rothstein, 1996; Fabiani & Friedman, 1995; Knight, 1984). Because this P3 is elicited under different conditions than the P3a, it has been labeled the ‘‘novelty P3’’ (e.g., Friedman, Simp-

SOUND FAMILIARITY AND NOVELTY ERP

33

son & Hamberger, 1993), as it is unclear, at the current state of knowledge, whether the P3a and the novelty P3 are identical. In the present investigation, the novel events were complex environmental sounds that varied in their degree of familiarity. At the end of the novelty oddball task each subject was asked to listen carefully to the sounds and to name them. Based on the Fabiani, Kazmerski, Cycowicz, and Friedman (1996) sound database, which includes modal names and conceptual names for 96 environmental sounds, the individual subject’s naming responses were classified into correct and incorrect conceptual names. The conceptual name was derived from the most common name given to a sound, but also included names related to it (e.g., synonyms, superordinates). Correct conceptual identifications were defined as familiar sounds, while incorrect conceptual identifications or ‘‘do not know’’ responses were defined as unfamiliar sounds. In previous research, these same environmental sounds were found to elicit two ERP components, the novelty P3 and the P32 (Cycowicz et al., 1996; Fabiani & Friedman, 1995; Friedman et al., 1993b; Kazmerski & Friedman, 1995). It has been suggested that the novelty P3 reflects a passive shift of attention, i.e., orienting (Friedman et al., 1993b). Previous studies have assessed the effect of novel repetition on the amplitude and scalp distribution of the novelty P3. Knight (1984) analyzed the repetition of a single novel event (a dog bark) and found a monotonic decrement in amplitude of the P3 component from the first to the fifth presentation of the novel stimuli, as well as a shift from a relatively frontally oriented to a relatively parietally oriented scalp topography. Similar results were reported by Courchesne (1978), who assessed the effect of recurrence of novel visual stimuli on the P3 component. In like fashion, Friedman and Simpson (1994) reported that novelty P3 amplitude decreased and its topography changed from a frontal to a posterior focus when sequentially presented unique novel sounds were used as eliciting stimuli. Furthermore, Cycowicz et al. (1996) and Kazmerski and Friedman (1995) demonstrated that the repetition of identical novel sounds also resulted in decreased novelty P3 amplitude and in a shift from a frontally to a parietally oriented topography. It has been suggested (e.g., Fabiani & Friedman, 1995; Knight, 1984) that the novelty P3 reflects the activity of a putative neuronal circuit that includes both anterior and posterior brain regions. In this view, the initial frontally oriented topography of the novelty P3 would reflect the involvement of the prefrontal cortex, perhaps indicating the formation of a working memory template. This memory template is maintained over time so that stimuli which are initially uncategorized are subsequently categorized into a distinct class of rare (i.e., ‘‘novel’’) events. Thus, repetition and recurrence of novel events (i.e., greater ‘‘experience’’) lead to a reduction in frontal activity, so that only the more posterior aspects of this presumed neural circuit predominate, resulting in a parietally oriented shift in scalp topography. This interpre-

34

CYCOWICZ AND FRIEDMAN

tation is consistent with the data and conclusion of Courchesne (1978), who suggested that stimuli for which no template exists elicit a frontally oriented P3, while stimuli that are categorizable (i.e., stimuli that match an existing template) elicit a more parietally oriented P3 scalp distribution. Based upon the results reviewed above, it was expected that repetition of both familiar and unfamiliar sounds would reduce the amplitude of the novelty P3. However, because familiar sounds may activate memory traces engendered by both repetition within the experiment and experience acquired outside the laboratory setting, a greater reduction for the familiar sounds than for the unfamiliar sounds was expected. Furthermore, since sound repetition alone causes a shift to a more parietally oriented topography, another expectation was that the most pronounced effect of repetition and familiarity would appear at the posterior scalp sites. The second ERP component, the P32, which is more posteriorly distributed than the novelty P3, has been interpreted as reflecting the subject’s attempt to categorize or linguistically encode (name) the novel events (Friedman et al., 1993b). With the repetition of novel sounds, P32 amplitude was enhanced, particularly at parietal sites (Cycowicz et al., 1996), and resembled the effect of word and picture repetition on P3b amplitude (Friedman, Hamberger, & Ritter, 1993; Rugg, Furda, & Lorist, 1988; see Rugg & Doyle, 1994 for a review). On this basis, we hypothesized that familiarity and repetition would interact in affecting P32 amplitude, such that unfamiliar repeated sounds would elicit a relatively larger increase in P32 amplitude than P32 amplitude elicited by repeated but familiar sounds. To summarize, the purpose of the present experiment was to investigate the modulation of the novelty P3 and P32 components as a function of the relative familiarity of the eliciting brief environmental sounds. The relative familiarity of words has been shown to affect both prelexical (Polich & Donchin, 1988) and postlexical (Rugg, 1990) stages of word processing. Assuming that information processing of both words and sounds share some similar mechanisms, the ERP components associated with the processing of environmental sounds should be modulated by sound familiarity. METHOD

Subjects Sixteen young adults (mean age 23.13 6 1.54 years; age range 22–28 years; 10 women and 6 men) were recruited for study by notices posted at the Columbia University Medical Center Complex. All subjects were native English speakers and received payment for their participation. Subjects with a history of head trauma, neurological disorders, learning disabilities, or hyperactivity were not admitted into the study. Subjects were administered the vocabulary and block design subtests from the Wechsler Adult Intelligence Scale–Revised (WAIS-R; Wechsler, 1981), and the modified Mini-Mental Status exam (mMMS, Mayeux, Stern, Rosen, & Leventhal, 1981), a measure of general cognitive function (maximum score 57). These tests were administered to identify subjects qualified to participate in the study,

SOUND FAMILIARITY AND NOVELTY ERP

35

and all subjects scored within normal limits. A socioeconomic status index was obtained based on educational level and occupation for each subject (Watt, 1976), and its mean was 52 6 3.22, indicating that subjects were of middle-class background. Informed consent was obtained from all subjects. Since the task involved listening to sounds, pure tone audiometry was administered. Hearing thresholds were tested at 250, 500, 1000, 2000, and 4000 Hz, and subjects met the following criteria: no more than a 35-dB mean hearing loss across frequencies; less than a 20-dB difference between ears at each frequency; less than a 30-dB difference between the best and the worst thresholds. The decibel level at which all stimuli were presented was adjusted for any subject who showed a mean hearing loss greater than 0 dB (averaged across frequencies and ears) by increasing the intensity of the stimuli (from 75 dB) by the mean dB hearing loss. Using this adjustment procedure, the mean intensity for both pure tones and environmental sounds was 79 dB (range 74–90).

Stimuli The stimuli were pure tones and environmental sounds. The pure tones were 500 Hz (high) and 350 Hz (low), each with a duration of 336 ms. The environmental sounds were 48 unique sounds (hereafter referred to as novels) that formed part of a larger corpus of environmental sounds described in detail by Fabiani et al. (1996). These sounds came from four sound categories: animal, human, musical instrument, and artificial or machine. The list of sounds appears in Appendix A. The novel sounds varied in duration from 159 to 399 ms (mean 5 336 6 61) and were matched for peak equivalent SPL to the pure tones using a decibel meter.

Procedure Subjects performed a standard auditory oddball task, a novelty oddball task, and a sound naming task. In the standard oddball task, subjects were asked to press a button when they heard rare tones (targets; p 5 .12) embedded in a series of frequent tones (standards; p 5 .88). Novelty oddball task. Following the standard oddball task, subjects were presented with 10 blocks of 80 trials each, with interstimulus intervals of 1275 ms. Environmental sounds (novel stimuli; p 5 .10) were randomly intermixed with standard (p 5 .80) and target (p 5 .10) tones. Subjects were not informed of the occurrence of the novel stimuli. When subjects asked about the novel sounds, they were instructed to continue responding to the target tones. There were 48 different nontonal, novel sounds. Thirty-two of them were presented twice, with each block containing 8 novel sounds. These 8 novel sounds were composed of two stimuli from each of the four different sound categories described above. In the first two blocks, all novel sounds were new, while in the rest of the blocks only half were new. The 16 novel sounds that did not repeat (unique) comprised half of the novel items in the first two and the last two blocks (4 in each block). Repetition of the novel stimuli occurred two blocks after their initial presentation, such that the novel stimuli initially presented in the first block, for example, were repeated in the third block. The novel events that repeated were rotated across blocks, so that across subjects each sound was presented equally often in each block. Stimuli were randomized separately for each subject, with the restrictions that a target or a novel could not occur as the first or the last stimulus and that two targets or novels could not be presented sequentially. The target ERP data from the two oddball tasks have been previously described, as have the novel ERP data with an emphasis on the effect of repetition (Cycowicz et al., 1996). Only the ERPs elicited by novel sounds during the novelty oddball task will be detailed here, averaged according to whether those sounds represented familiar or unfamiliar sound concepts.

36

CYCOWICZ AND FRIEDMAN

Sound naming task. After subjects completed the novelty oddball task, they performed a sound naming task. In that task, each subject was presented with a different random ordering of the environmental sounds one at a time. Subjects were instructed to listen to each sound, to name each as accurately as possible, and to state ‘‘I do not know’’ if they were unable to identify a sound. All subjects were asked to name the 48 sounds that had been presented during the novelty oddball blocks. In addition, 8 subjects were asked to name 96 sounds, including 48 sounds that had not been heard during the novelty oddball task. The experimenter transcribed verbatim each response before initiating the delivery of the next sound.

EEG Recordings EEG and EOG were recorded continuously with a bandpass of 0.01 to 30 Hz (5.3 s time constant) and were digitized at a rate of 200 Hz. EEG was recorded using an Electrocap with placements at Fz, Cz, Pz, F3, F7, C3, P3, T5, F4, F8, C4, P4, and T6. All leads were referred to nosetip. Vertical EOG was recorded bipolarly from electrodes placed on the supraorbital and infraorbital ridges of the right eye, and horizontal EOG was recorded bipolarly from electrodes placed on the outer canthi of the two eyes. ERP data were epoched off-line for 100-ms pre- and 1175-ms poststimulus periods. Eye movement artifacts were corrected offline by means of a procedure developed by Gratton, Coles, and Donchin (1983). In addition, the single trials were visually inspected and trials containing muscular or other artifacts were marked and excluded from further analysis.

Data Analyses For the purpose of the present study, only the 32 sounds that repeated were analyzed. For each sound, the sound database from Fabiani et al. (1996) was used to determine the ‘‘correctness’’ of the names provided by the 16 subjects. Fabiani et al. (1996) provided a modal name (the name given by most subjects) as well as a concept agreement index for each sound. The modal concept was derived from the modal name, but it also included names related to each sound (e.g., synonyms, superordinates). In the present study, the modal concept was used to assess individual subject’s naming responses. A correct name was defined as a name that was included in the conceptual coding provided by the Fabiani et al. database. For example, naming the call of a cardinal a ‘‘nightingale’’ was considered correct because the conceptual name of this sound was a bird call. Sounds that were named within this conceptual schema were defined as familiar sounds. Names that did not agree with the conceptual definition of the sound (e.g., the sound of a baby crying named a ‘‘car’’), as well as ‘‘do not know’’ responses, were treated as incorrect names and were categorized as unfamiliar sounds. For each subject, the naming responses were coded as familiar or unfamiliar based on his or her own responses. These codings were used to average each subject’s ERP as a function of his or her own sound familiarity judgements. For the ERP data, averaged voltage measurements were computed separately for the first (novel 1) and second (i.e., repeated, novel 2) presentations of the novel sounds, were averaged according to whether they were elicited by familiar or unfamiliar sounds. Two ERP components that were defined on the basis of previous studies with this paradigm (Cycowicz et al., 1996; Kazmerski & Friedman, 1995) were measured: novelty P3 within a 220- to 400-ms latency window, and P32 within a 450- to 800-ms latency window. Differences in scalp distribution across conditions were inferred from the interaction of condition and electrode site, after the data had been normalized using the root mean square method described by McCarthy and Wood (1985). The root mean square was applied using the averaged voltages based on the specific time windows of the novelty P3 (220–400 ms) and P32 (450–800 ms) components. This technique removes overall amplitude differences between conditions, so that only shape is considered in the analysis of variance on the normal-

SOUND FAMILIARITY AND NOVELTY ERP

37

ized data. This enables unambiguous interpretation of scalp distribution differences when a significant condition by electrode site interaction is found after normalization. Analyses of variance (ANOVA) were performed using the BMDP-4V program (Dixon, 1987). The Greenhouse–Geisser epsilon correction (Jennings & Wood, 1976) was used where appropriate. Uncorrected degrees of freedom are reported below along with the epsilon value (ε); the p values reflect the epsilon correction. Where appropriate, significant main effects and interactions were followed with simple effects procedures and/or post hoc analyses using the Tukey Honestly Significant Difference (HSD) test.

RESULTS

Behavioral Data Novelty oddball task. Subjects were very accurate in detecting the target (99.45%) and very occasionally made false alarms to standard (0.01%) and to novel (2.58%) presentations. Mean reaction time (RT) for correct target detections was 460 6 70 ms. Sound naming task. The average number of correctly named (i.e., familiar) novel sounds was 21.38 (range 15–26) of 32 repeated sounds. Across the four categories of environmental sounds, the distribution of correctly named sounds was as follows: animal 5.88, human 5.94, musical instrument 4.38, artificial or machine sounds 5.19. These data were submitted to a repeatedmeasures ANOVA, which revealed a significant difference (F(3, 45) 5 5.12, p , .004) in sound familiarity across the four categories. A post hoc analysis revealed that correct naming was significantly poorer for musical instrument sounds than for animal and human sounds, but did not differ from correct naming for artificial sounds. ERP Waveforms Figures 1 and 2 depict the grand mean ERP waveforms recorded at the scalp midline elicited by first (novel 1) and second (novel 2) presentations of the novel sounds. In Fig. 1, the ERPs elicited by the familiar and unfamiliar sounds are superimposed for novel 1 and novel 2 separately, while in Fig. 2, the ERPs to novel 1 and novel 2 are superimposed separately for the familiar and the unfamiliar sounds.1 A few phenomena worth noting can be seen by inspection of Figs. 1 and 2. First, the ERPs elicited by the novels are composed of two late positive deflections—the novelty P3 and the P32 —consistent with previous research (e.g., Fabiani & Friedman, 1995; Friedman & Simpson, 1994). Second, as is evident from inspection of Fig. 1, the novelty P3 amplitudes elicited by 1 The mean numbers of trials (and range) for each condition were as follows: novel 1 familiar sounds, 19 (14–24); novel 1 unfamiliar sounds, 10 (6–17); novel 2 familiar sounds, 20 (15– 28); novel 2 unfamiliar, 10 (5–17). Although there were fewer unfamiliar sounds than familiar sounds, the signal-to-noise ratios for all conditions were quite good, and for each individual’s set of waveforms the two components (novelty P3 and P32) could be easily identified.

38

CYCOWICZ AND FRIEDMAN

FIG. 1. Grand mean ERP waveforms elicited by first (novel 1) and second (novel 2) presentations of novel stimuli as a function of familiarity at the scalp midline (Fz, Cz, Pz). Arrows mark stimulus onset with time lines every 100 ms.

FIG. 2. Grand mean ERP waveforms elicited by familiar and unfamiliar novel stimuli as a function of repetition at midline electrode sites. Arrows mark stimulus onset with time lines every 100 ms.

39

SOUND FAMILIARITY AND NOVELTY ERP

TABLE 1 Midline ANOVA Results for Novel Repetition and Sound Familiarity in the Novelty Oddball Task Novelty P3

P32

Factor

F

df

ε

p

F

df

ε

p

Repetition Familiarity Electrode Repetition 3 Familiarity Repetition 3 Electrode Familiarity 3 Electrode Repetition 3 Familiarity 3 Electrode

0.82 2.10 16.43 6.73 4.43 2.39 3.05

1/15 1/15 2/30 1/15 2/30 2/30 2/30

— — .78 — .84 .65 .82

n.s. n.s. .000 .020 .028 n.s. .070

0.76 0.01 23.01 1.63 11.42 0.09 0.15

1/15 1/15 2/30 1/15 2/30 2/30 2/30

— — .69 — .69 .67 .62

n.s. n.s. .000 n.s. .001 n.s. n.s.

Note. n.s., not significant.

familiar and unfamiliar sounds do not differ for the first presentation (novel 1), but do differ for the second presentation (novel 2), with the novelty P3 elicited by the unfamiliar sounds showing greater amplitude. Third, as can be seen in Fig. 2, the effect of repetition is different for familiar and unfamiliar sounds. For the familiar events, repetition leads to a decrease in novelty P3 amplitude, which is greatest at the Fz and Cz locations. By contrast, for the unfamiliar sounds, repetition leads to an increase in amplitude, but only at the posterior scalp location. As can be observed in Fig. 1, the ERPs elicited by familiar sounds exhibit larger P32 amplitudes than the ERPs elicited by unfamiliar sounds for novel 1, whereas for novel 2, familiar sounds display somewhat smaller amplitudes than unfamiliar sounds. Inspection of Fig. 2 shows that only unfamiliar sounds appear to produce a consistent difference in P32 amplitude between novel 1 and novel 2, with novel 2 eliciting larger amplitude than novel 1. None of these effects appears to occur for the negativity preceding the novelty P3. Raw Amplitude Analyses To corroborate statistically these visual impressions, the averaged voltage data of the novelty P3 and the P32 were submitted to repeated-measures ANOVAs with Repetition (novel 1, novel 2), Familiarity (familiar, unfamiliar), and Electrode Site (Fz, Cz, Pz) as within-subjects factors. The results of these amplitude analyses are presented in Table 1. The averaged voltage data of the negativity preceding the novelty P3 were also measured and submitted to a similar ANOVA, but none of the main effects or interactions was significant ( ps . .05). Novelty P3. As can be seen in Table 1, for the novelty P3, aside from Electrode Site, neither main effect was reliable. However, both the Repetition by Electrode Site and the Repetition by Familiarity interactions were reliable.

40

CYCOWICZ AND FRIEDMAN

TABLE 2 Thirteen Electrode Site ANOVA Results for Novel Repetition and Sound Familiarity in the Novelty Oddball Task Novelty P3

P32

Factor

F

df

ε

p

F

df

ε

p

Repetition Familiarity Electrode Repetition 3 Familiarity Repetition 3 Electrode Familiarity 3 Electrode Repetition 3 Familiarity 3 Electrode

0.01 0.51 21.71 9.04 1.70 0.87 5.42

1/15 1/15 12/180 1/15 12/180 12/180 12/180

— — .21 — .27 .19 .25

n.s. n.s. .000 .009 n.s. n.s. .003

2.54 0.03 19.85 2.60 5.64 0.44 2.20

1/15 1/15 12/180 1/15 12/180 12/180 12/180

— — .18 — .29 .28 .19

n.s. n.s. .000 n.s. .001 n.s. n.s.

Note. n.s., not significant.

The Repetition by Electrode Site interaction replicates previous findings (Cycowicz et al., 1996; Kazmerski & Friedman, 1995). In brief, it reflects the fact that novel repetition induced a reduction of amplitude at the frontal site, which was not present at the Pz location. This interaction remained after normalization (F(2, 30) 5 4.32, p , .03, ε 5 0.80), indicating that the scalp distribution of the novelty P3 changed from frontally to more posteriorly oriented from the first to the second presentation as previously described (Cycowicz et al., 1996; Kazmerski & Friedman, 1995). The important interaction of Repetition and Familiarity was examined by tests for simple effects. These indicated that for novel 1, novelty P3 amplitude did not differ for familiar and unfamiliar sounds (F(1, 15) 5 .21, p . .05), whereas for novel 2, novelty P3 amplitude was significantly larger for unfamiliar than for familiar sounds (F(1, 15) 5 11.94, p , .004). In addition, there was a significant difference between novel 1 and novel 2 for familiar sounds (F(1, 15) 5 5.03, p , .04), such that repetition elicited a decrease in novelty P3 amplitude. There was no such difference for unfamiliar sounds as a function of repetition (F(1, 15) 5 1.46, p . .2). Similar results, with one important addition, were obtained when the analysis was performed on the data from the full complement of 13 electrode sites (Table 2). The interaction of Repetition and Familiarity was significant, and simple effects tests again revealed that only the repeated novels elicited significantly larger novelty P3 amplitudes for unfamiliar sounds than for familiar sounds (p , .007). Furthermore, post hoc tests revealed that the repetition effect (decreased amplitude) was significant only for the familiar sounds (p , .05). However, the important interaction of Repetition, Familiarity, and Electrode Site, which was not reliable in the midline analysis, was significant. The raw amplitude values for the novelty P3 at the 13 electrode sites for familiar and unfamiliar sounds and for the first and second presentations

SOUND FAMILIARITY AND NOVELTY ERP

41

FIG. 3. Grand mean raw amplitude values for the novelty P3 elicited by novel 1 and novel 2, and familiar and unfamiliar sounds at the 13 electrode sites. These data depict the triple interaction of Repetition, Familiarity, and Electrode Site. A significant difference at p , .05 between the contrasted conditions is indicated by an asterisk.

are depicted in Fig. 3. Simple effects and post hoc tests ( ps , .05, asterisks in Fig. 3) revealed that repetition of familiar sounds elicited a decrease in amplitude primarily at the fronto-central electrode sites (Fig. 3, top left), while repetition of unfamiliar sounds elicited an increase in amplitude at the posterior electrode locations with no differences at frontal sites (Fig. 3, top right). In addition, while there was no amplitude difference between familiar and unfamiliar sounds for first presentation novels at any electrode location (Fig. 3, bottom left) larger novelty P3 amplitudes were elicited by unfamiliar sounds than by familiar sounds at most electrode sites for the second presentation of novel events (Fig. 3, bottom right). These findings for the raw ampli-

42

CYCOWICZ AND FRIEDMAN

tudes are consistent with the existence of multiple generators subserving the novelty P3. Therefore, this interaction is dealt with under Scalp Topography Analyses, where the data are normalized to allow unequivocal interpretation. P32. The main effect of Electrode Site was significant, but it was modulated by the interaction of Electrode Site and Repetition for both the midline and the 13 electrode site analyses (see Tables 1 and 2). Similar to previous findings (Cycowicz et al., 1996; Kazmerski & Friedman, 1995), simple effects tests showed that repetition induced an increase in P32 amplitude at the Pz location (or other posterior electrode sites in the 13 site analysis), but not at Fz and Cz (or other centro-frontal sites in the 13 site analysis). The interactions of Repetition and Familiarity, and of Repetition, Familiarity, and Electrode Site, were not significant for the P32 component, in either the midline or the 13 electrode site analyses (see Tables 1 and 2). Scalp Topography Analyses (13 Site Analyses) Because the interaction of Repetition, Familiarity, and Electrode Site was significant for the novelty P3 component in the raw data, the data from the 13 electrode sites were normalized with respect to both familiarity and repetition. This was done by using the novel 1 familiar condition as the scalar and normalizing the three remaining conditions (novel 1 unfamiliar, novel 2 familiar, and novel 2 unfamiliar) with respect to this scalar. Novelty P3. To assess changes in scalp distribution as a function of sound familiarity and repetition, the normalized data were submitted to a repeatedmeasures ANOVA. For the novelty P3, the three-way interaction of Repetition, Familiarity, and Electrode Site was significant (F(12, 180) 5 4.08, p , .01, ε 5 0.27). The normalized amplitude values for the novelty P3 at the 13 electrode sites for familiar and unfamiliar sounds and for the first and second presentations are depicted in Fig. 4. Post hoc tests ( ps , .05) revealed that repetition of the familiar sounds led to a relative increase in amplitude at the Pz site (Fig. 4, top left). By contrast, repetition of the unfamiliar sounds elicited a relative amplitude increase at posterior sites (P4, T5, and T6), as well as a relative decrease in amplitude at the Fz and Cz locations (i.e., a change to a more posterior topography; Fig. 4, top right). In addition (Fig. 4, bottom left), post hoc tests revealed that at the posterior sites (T6 and P4) relatively larger amplitudes were elicited by familiar sounds than by unfamiliar sounds for the first presentation. However, for the second presentation, only at Cz larger amplitude was elicited by familiar sounds than by unfamiliar sounds. At the posterior temporal sites (T5 and T6), amplitudes were relatively larger for the unfamiliar sounds (Fig. 4, bottom right). In summary, the processing of novel sounds as indexed by the novelty P3 is modulated by both repetition and familiarity. The amplitude of novelty P3 elicited by the first presentation of familiar and unfamiliar sounds, did

SOUND FAMILIARITY AND NOVELTY ERP

43

FIG. 4. Grand mean normalized amplitude values for the novelty P3 elicited by novel 1 and novel 2, and familiar and unfamiliar sounds at the 13 electrode sites. These data depict the triple interaction of Repetition, Familiarity, and Electrode Site. A significant difference at p , .05 between the contrasted conditions is indicated by an asterisk.

not differ. However, novelty P3 amplitude elicited by repeated familiar sounds was smaller than novelty P3 elicited by repeated unfamiliar sounds. Moreover, while repetition of unfamiliar sounds elicited amplitude increases at posterior electrode sites, repetition of familiar sounds elicited amplitude decreases at frontal electrode sites, suggesting a functional dissociation of the frontal and posterior aspects of the novelty P3. This dissociation was further supported by changes in scalp topography as a function of both repetition and familiarity.

44

CYCOWICZ AND FRIEDMAN

DISCUSSION

The results presented in this paper demonstrate that the relative familiarity of environmental sounds modulates the amplitude and topography of the novelty P3 in a more complex way than was anticipated. Repetition of the novel sounds elicited a significant reduction of novelty P3 amplitude over fronto-central scalp sites for familiar sounds, whereas for unfamiliar sounds repetition induced an increase in amplitude at primarily posterior scalp locations (Fig. 3). Moreover, repetition of the unfamiliar sounds elicited a greater change in scalp topography (a change from frontally focused activity to more posteriorly focused activity) than repetition of the familiar sounds (Fig. 4). These findings, however, must be considered tentative as the number of trials entering the ERP averages for both familiar and unfamiliar novel events was small. Thus, the current findings are in need of confirmation with a novelty oddball paradigm that allows for a greater number of trials in the novel 1 and 2 familiar and unfamiliar categories. Previous researchers have shown a decrease in novelty P3 amplitude with recurrence of novel events whether the eliciting novel events were a sequence of unique novel sounds (Friedman & Simpson, 1994), unique visual novel stimuli (Courchesne, 1978), repetitions of a single novel event (Knight, 1984), or repetition of identical novel sounds (Cycowicz et al., 1996; Kazmerski & Friedman, 1995). These studies treated all novel stimuli as belonging to a single class and thus assessed novelty P3 amplitude only as a function of the temporal order in which the stimuli were presented (i.e., recurrence and/or repetition). The present investigation introduces subjective familiarity of the stimulus as an additional variable, assessing the effect of novel stimulus familiarity in interaction with repetition on ERP amplitude and scalp topography. Taking this additional variable into account, these data show that the degree of novelty P3 amplitude change depends upon the relative familiarity of the repeating novel event, and that sound familiarity appears to differentially affect the various generators that give rise to the novelty P3. Considerable caution must be exercised when inferring, based on scalprecorded data alone, the origin of electrical activity within the brain. It is possible, however, based on findings from studies of patients with welllocalized brain lesions, on intracranial recordings from humans, and source modeling in normal human subjects, together with distributional data from scalp recordings, to obtain more informed estimates of where in the brain the generators are likely to be located. In the following paragraphs we will describe the familiarity effect and its possible neuronal underpinnings. A variety of evidence from experiments with animals and patients with focal brain lesions suggests that the prefrontal cortex is important for processing external events prior to long-term memory encoding. Humans with lesions in the dorsolateral prefrontal cortex exhibit abnormalities in performing a variety of cognitive tasks, including the detection of novel events,

SOUND FAMILIARITY AND NOVELTY ERP

45

and demonstrate a lack of habituation to repetitive novel stimuli (Glaser & Griffin, 1962; Knight, 1984; Woods & Knight, 1986). Prefrontal damage also results in a markedly reduced P3 to novel stimuli, whether the novel stimuli are auditory (Knight, 1984) or visual (Knight, 1990). In these patients, the reduced P3 elicited by novels is also accompanied by maximum amplitude at the parietal midline site, unlike the P3 recorded in the control group, which shows a frontally oriented distribution (Knight, 1984). Patients with focal damage centered in the temporo-parietal junction, who also show reduced orienting to distracting stimuli (Lamb, Roberston, & Knight, 1989), provide evidence of another brain region presumably involved in generating the novelty P3. Although these patients performed an auditory novelty oddball task as accurately as subjects in the control group, they did not produce a novelty P3 with a scalp distribution similar to that of controls (Knight, Scabini, Woods, & Clayworth, 1989). The lesions abolished the P3 over the central and parietal scalp (its voltage was reduced to noise level), but frontal activity was partially preserved. Another difference between frontal and temporo-parietal brain lesions is in the extent to which they affected the topography of the novelty P3. While unilateral prefrontal damage reduced P3 amplitudes over the lesioned hemisphere only, unilateral damage to the temporo-parietal junction abolished the P3 over the parietal scalp of both hemispheres (Knight, Grabowecky, & Scabini, 1995; Knight et al., 1989). Additional evidence for the involvement of anterior and posterior neuronal generators in processing novel stimuli has been demonstrated in a series of studies in which depth electrodes were implanted to localize seizure origin prior to surgical treatment of intractable epilepsy (Halgren, Baudena, Clarke, Heit, Liegeois, Chauvel, & Musolino, 1995a; Halgren, Baudena, Clarke, Heit, Marinkovic, Devaux, Vignal, & Biraben, 1995b; Baudena, Halgren, Heit, & Clarke, 1995). In these studies, intracerebral potentials were recorded while subjects performed an auditory oddball task with distractors (novels). The distractors elicited an early P3 component, labeled P3a by these investigators, with generators (indicated by local polarity reversals) in both the frontal lobes and the posterior cortex (Baudena et al., 1995). In addition, using source modeling techniques to localize neuronal activity (e.g., Scherg, 1990), Simpson, Fabiani, and Friedman (submitted for publication) also reported the involvement of posterior and frontal cortical regions in producing the novelty P3 in healthy young and older adults. It has been suggested that the frontal aspect of the novelty P3 is involved in the formation and/or maintenance of working memory templates (Fabiani & Friedman, 1995) or in attempting to classify the unexpected event (Friedman & Simpson, 1994). Thus the frontal lobes appear to play an essential role in the detection of novel events. In contrast, it has been suggested that the posterior aspect of the novelty P3 reflects neuronal processes involved in matching sensory information to memory templates (Knight et al., 1995).

46

CYCOWICZ AND FRIEDMAN

These brain regions have reciprocal connections, and they also interact with other brain areas involved in memory processing, such as the hippocampal formation (Goldman-Rakic, Selmon, & Schwartz, 1984). The effect of sound familiarity described in the present paper emerged only after repetition, as there was no difference in novelty P3 amplitude between familiar and unfamiliar sounds on their first presentation. This implies that on first presentation both familiar and unfamiliar sounds were similarly detected as novel events and presumably similarly activated putative frontal lobe generators. However, second presentations of the novel sounds elicited differential activation of the frontal and posterior aspects of the novelty P3 as a function of sound familiarity. There are several findings that support this view. First, based on the raw amplitudes, an overall smaller novelty P3 amplitude was found for repeated familiar sounds than for repeated unfamiliar sounds. Second, for unfamiliar sounds, the difference in novelty P3 amplitude between novel 1 and novel 2 emerged predominantly at the posterior locations with no differences observed at frontal sites; at these posterior locations, relative to novel 1, repeated novels elicited increased novelty P3 amplitude. By contrast, for familiar sounds, the difference in novelty P3 amplitude between novel 1 and novel 2 emerged predominantly at the frontal locations; at these frontal locations, relative to novel 1, repeated novels elicited decreased novelty P3 amplitude. Third, these amplitude changes observed in the raw data for the unfamiliar sounds were manifested topographically (i.e., after normalization) as a relative decrease in frontal activity and a relative increase in posterior activity. However, for familiar repeated sounds, the change in topography stemmed from a relative increase in activity at the parietal site only. These results provide additional evidence for the functional dissociation of the anterior and posterior aspects of the scalp recorded novelty P3. Furthermore, they also suggest that these two aspects of the novelty P3 are most likely subserved by unique neuronal generators. That novelty P3 amplitude was smaller at all electrode locations for familiar than for unfamiliar repeated sounds suggests that both the anterior and the posterior aspects of the novelty P3 are sensitive to the degree of novelty within the context of the experiment. However, the frontal aspect of the novelty P3, which is assumed to be involved in the maintenance and/or formation of memory templates or in the classifying of novel sounds (Fabiani & Friedman, 1995; Friedman & Simpson, 1994), exhibited a reduction in amplitude compared to the posterior aspect for familiar sounds only. This can be explained by the fact that for familiar sounds a memory template may exist, thus requiring less processing on second presentation. On the other hand, with repetition of unfamiliar sounds, the posterior aspect, which is assumed to be involved in the matching of sensory information to memory templates (Knight et al., 1995), reflects greater processing due to the relative unfamiliarity of the eliciting sounds.

SOUND FAMILIARITY AND NOVELTY ERP

47

As postulated earlier, the longer-latency P32 component may reflect an attempt to categorize and/or name the stimulus that had just been presented (Friedman et al., 1993b). The result of such an attempt may provide information for that aspect of the presumed neuronal network responsible for extracting features critical for stimulus detection. This aspect of the novelty P3 system reflected by the P32 may modify the memory traces of familiar and unfamiliar sounds differently, as evidenced by the differential effect of familiarity on the frontal and posterior aspects of the novelty P3. Thus, the novelty P3 could reflect the activity of a novelty detection system, displaying special sensitivity to the repetition of less familiar events. In a different experimental paradigm, which involved recognition memory for high- and low-frequency words, Rugg and Doyle (1992) found that P600 amplitude was affected by repetition of low-frequency words only. Rugg and Doyle (1992) suggested that the low- and high-frequency words received differential processing as a function of repetition. Although the present research involved different stimuli and a quite different task, its results also suggest differential processing of familiar and unfamiliar events as a function of repetition. From an evolutionary point of view, it is advantageous to have a brain network that can distinguish not only between familiar and novel events, but also between events that vary in their degree of novelty. The second P3 component, the P32, was expected to show a repetition priming effect similar to that found with pictures and words (e.g., Friedman et al., 1993b; Rugg & Doyle, 1994), namely an increase in amplitude. Moreover, a larger increase was expected for the unfamiliar sounds than for the familiar sounds based on earlier findings that described a larger repetition priming effect for low-frequency words than for high-frequency words (Rugg, 1990). This expectation was not supported, as the amplitude difference among the conditions did not reach statistical significance. However, visual inspection of the waveforms (see Figs. 1 and 2) suggests that the direction of P32 activity was as predicted. The lack of statistical significance may have been due to the relatively small number of trials in the ERP averages and/or to the large variability in naming performance among subjects, i.e., to difficulties in naming and identifying brief environmental sounds. Unlike picture naming of concrete objects, which produces naming responses that are relatively consistent across subjects, the naming responses of the brief sounds used here varied greatly (see Fabiani et al., 1996). In conclusion, the data of the present study suggest that different stages in the processing of novel information can be detected by measuring brain electrical activity. It is possible that similar types of mechanisms are involved in processing words and environmental sounds, because both kinds of stimuli exhibited differential brain activity as a function of stimulus familiarity. The novelty P3 component appears to involve multiple generators, some of which presumably originate in anterior brain regions and may reflect orienting to and classification of novel events. Other, more posterior generators may re-

48

CYCOWICZ AND FRIEDMAN

flect a process that distinguishes among different degrees of novelty or familiarity. Semantic analysis of these novel events may be indexed by the P32 component, which is elicited some 200–300 ms after the novelty P3. The data suggest that P32 may also be modulated by sound familiarity, but further research is needed to clarify this effect. APPENDIX A

The 48 novel sounds used in the present study for each sound category are listed. In parentheses is the number of subjects who correctly identified the sounds. Sounds that were presented only once and that were not included in the analysis are marked with an asterisk. Animal sounds. Cat (12), Cricket (10)*, Frog (15), Horse (6)*, Mosquito (13), Whale (6), Cow (13), Dog (16), Wolf (7), Owl (6)*, Cardinal (12), Mallard (15)*. Human sounds. Baby Crying (9), Burp (15), Clapping (11), Coughing (16), Hiccup (15), Kiss (2)*, Laughing (16), Sneeze (11)*, Speech-I (5), Speech-II (15)*, Throat Clearing (11)*, Whistle (8). Musical instruments. Bugle (16), Cymbal (7), Fanfare (12)*, Piano Notes (6), Guitar (7), Notes (14)*, Piccolo (5), Synthesizer (10)*, Synthesizer (13), Oboe (11), Xylophone (5), Drum (9)*. Noise and artificial sounds. Car Starting (5), Cuckoo Clock (14), Car Horn (15), Knock (12)*, Water (5)*, Phone (11), Boing (11)*, Video Game (7), Machine Gun (16), Ahooah (4), Ooahoo (9)*, Zap Down (11). REFERENCES Ballas, J. A. 1993. Common factors in the identification of an assortment of brief everyday sounds. Journal of Experimental Psychology: Human Perception and Performance, 19, 250–267. Baudena, P., Halgren, E., Heit, G., & Clarke, J. M. 1995. Intracerebral potentials to rare target and distractor auditory and visual stimuli. III. Frontal cortex. Electroencephalography and Clinical Neurophysiology, 94, 251–264. Courchesne, E. 1978. Changes in P3 waves with event repetition: Long-term effects on scalp distribution and amplitude. Electroencephalography and Clinical Neurophysiology, 45, 468–482. Courchesne, E., Hillyard, S. A., & Galambos, R. 1975. Stimulus novelty, task relevance, and the visual evoked potential in man. Electroencephalography and Clinical Neurophysiology, 39, 131–143. Cycowicz, Y. M., Friedman, D., & Rothstein, M. 1996. An ERP developmental study of repetition priming by auditory novel stimuli. Psychophysiology, 33, 680–690. Dixon, W. D. 1987. The BMDP biomedical computer programs. Los Angeles: Univ. of California Press. Duncan-Johnson, C. C., & Kopell, B. 1981. The Stroop effect: Brain potentials localize the source of interference. Science, 214, 938–940. Fabiani, M., & Friedman, D. 1995. Changes in brain activity patterns in aging: The novelty oddball. Psychophysiology, 32, 579–594.

SOUND FAMILIARITY AND NOVELTY ERP

49

Fabiani, M., Kazmerski, V. A., Cycowicz, Y. M., & Friedman, D. 1996. Naming norms for brief environmental sounds. Psychophysiology, 33, 462–475. Forster, K. I., & Chambers, S. M. 1973. Lexical access and naming time. Journal of Verbal Learning and Verbal Behavior, 12, 627–635. Forster, K. I., & Davis, C. 1984. Repetition priming and frequency attenuation in lexical access. Journal of Experimental Psychology: Learning, Memory, and Cognition, 10, 680–698. Friedman, D. 1990. ERPs during continuous recognition memory for words. Biological Psychology, 30, 61–87. Friedman, D., & Simpson, G. V. 1994. ERP amplitude and scalp distribution to target and novel events: Effect of temporal order in young, middle-aged and older adults. Cognitive Brain Research, 2, 49–63. Friedman, D., Hamberger, M., & Ritter, W. 1993a. Event-related potentials as indicators of repetition priming in young and elderly adults: Amplitude, duration and scalp distribution. Psychology and Aging, 8, 120–125. Friedman, D., Simpson, G., & Hamberger, M. 1993b. Age-related changes in scalp topography to novel and target stimuli. Psychophysiology, 30, 383–396. Glaser, E. M., & Griffin, J. P. 1962. Influence of cerebral cortex on habituation. Journal of Physiology, 1960, 429–445. Goldman-Rakic, P. S., Selmon, L. D., & Schwartz, M. L. 1984. Dual pathway connecting the dorsolateral prefrontal cortex with the hippocampal formation and parahippocampal cortex in the Rhesus monkey. Neuroscience, 12, 719–743. Gratton, G., Coles, M. G. H., & Donchin, E. 1983. A new method for off-line removal of ocular artifact. Electroencephalography and Clinical Neurophysiology, 55, 468–484. Halgren, E., Baudena, P., Clarke, J. M., Heit, G., Liegeois, C., Chauvel, P., & Musolino, A. 1995a. Intracerebral potentials to rare target and distractor auditory and visual stimuli. I. Superior temporal plane and parietal lobe. Electroencephalography and Clinical Neurophysiology, 94, 191–220. Halgren, E., Baudena, P., Clarke, J. M., Heit, G., Marinkovic, K., Devaux, B., Vignal, J.-P., & Biraben, A. 1995b. Intracerebral potentials to rare target and distractor auditory and visual stimuli. II. Medial, lateral and posterior temporal lobe. Electroencephalography and Clinical Neurophysiology, 94, 229–250. Halgren, E., & Smith, M. E. 1987. Cognitive evoked potentials as modulatory processes in human memory formation and retrieval. Human Neurobiology, 6, 129–139. Jacoby, L. L., & Dallas, M. 1981. On the relationship between autobiographical memory and perceptual learning. Journal of Experimental Psychology: General, 110, 306–340. Jennings, J. R., & Wood, C. C. 1976. The E-adjustment procedure for repeated-measure analyses of variance. Psychophysiology, 13, 277–278. Johnson, R., Jr. 1986. A triarchic model of P300 amplitude. Psychophysiology, 23, 367–384. Johnson, R., Jr. 1993. On the neural generators of the P300 component of the event related potential. Psychophysiology, 30, 90–97. Kazmerski, V. A., & Friedman, D. 1995. Repetition of novel stimuli in an ERP oddball paradigm: Aging effect. Journal of Psychophysiology, 9, 298–311. Knight, R. T. 1984. Decreased response to novel stimuli after prefrontal lesions in man. Electroencephalography and Clinical Neuropsychology, 59, 9–20. Knight, R. T. 1990. Neural mechanisms of event-related potentials: Evidence from human lesion studies. In J. W. Rohrbaugh, R. Parasuraman, & R. Johnson, Jr. (Eds.), Eventrelated brain potentials: Basic issues and applications. New York: Oxford Univ. Press.

50

CYCOWICZ AND FRIEDMAN

Knight, R. T. 1994. Attention regulation and human prefrontal cortex. In A. M. Thierry et al. (Eds.), Motor and cognitive functions of the prefrontal cortex. Berlin: Springer Verlag. Knight, R. T., Grabowecky, M. F., & Scabini, D. 1995. Role of human prefrontal cortex in attention control. In H. H. Jasper, S. Riggio, & P. S. Goldman-Rakic (Eds.), Epilepsy and the functional anatomy of the frontal lobe. New York: Raven Press. Knight, R. T., Scabini, D., Woods, D. L., & Clayworth, C. C. 1989. Contributions of temporalparietal junction to the human auditory P3. Brain Research, 502, 109–116. Kutas, M., & Hillyard, S. A. 1984. Brain potentials during reading reflect word expectancy and semantic association. Nature, 307, 161–163. Lamb, M. R., Roberston, L. C., & Knight, R. T. 1989. Attention and interference in the processing of global and local information: Effects of unilateral temporal-parietal lesions. Neuropsychologia, 27, 471–483. Mayeux, R., Stern, Y., Rosen, J., & Leventhal, J. 1981. Depression, intellectual impairment, and Parkinson’s disease. Neurology, 31, 645–650. McCarthy, G., & Donchin, E. 1981. A metric thought: A comparison of P300 latency and reaction time. Science, 211, 77–80. McCarthy, G., & Wood, C. C. 1985. Scalp distributions of event-related potentials: An ambiguity associated with analysis of variance models. Electroencephalography and Clinical Neurophysiology, 62, 203–208. Nagy, M. E., & Rugg, M. D. 1989. Modulation of event-related potentials by words repetition: The effects of inter-item lag. Psychophysiology, 26, 431–436. Picton, T. W., & Hillyard, S. A. 1988. Endogenous event-related potentials. In T. W. Picton (Ed.), Human event-related potentials, EEG handbook. (Vol. 3) Pp. 361–425. Amsterdam/New York: Elsevier. Polich, J., & Donchin, E. 1988. P300 and the word frequency effect. Electroencephalography and Clinical Neurophysiology, 70, 33–45. Rugg, M. D. 1987. Dissociation of semantic priming, word and non-word repetition effects by event-related potentials. The Quarterly Journal of Experimental Psychology, 39A, 123–148. Rugg, M. D. 1990. Event-related brain potentials dissociate repetition effects of high- and low-frequency words. Memory and Cognition, 18, 367–379. Rugg, M. D., & Doyle, M. C. 1992. Event-related potentials and recognition memory for lowand high-frequency words. Journal of Cognitive Neuroscience, 4, 69–79. Rugg, M. D., & Doyle, M. C. 1994. Event-related potentials and stimulus repetition in direct and indirect tests of memory. In H. Heinz, T. Munte, & G. R. Mangun (Eds.), Cognitive electrophysiology. Cambridge, MA: Birkha¨user. Rugg, M. D., Furda, J., & Lorist, M. 1988. The effect of task on the modulation of eventrelated potentials by word repetition. Psychophysiology, 25, 55–63. Scarborough, D. L., Cortese, C., & Scarborough, H. 1977. Frequency and repetition effects in lexical memory. Journal of Experimental Psychology: Human Perception and Performance, 3, 1–17. Scherg, M. 1990. Fundamentals of dipole source analysis. In M. Hoke et al. (Eds.), Auditory evoked magnetic fields and potentials. Basel: Karger. Pp. 40–69. Squires, N. K., Squires, K. C., & Hillyard, S. A. 1975. Two varieties of long-latency positive waves evoked by unpredictable auditory stimuli in man. Electroencephalography and Clinical Neurophysiology, 38, 387–410. Stuart, G. P., & Jones, D. M. 1995. Priming the identification of environmental sounds. The Quarterly Journal of Experimental Psychology, 48a, 741–761.

SOUND FAMILIARITY AND NOVELTY ERP

51

Sutton, S., Braren, M., Zubin, J., & John, E. R. 1965. Evoked potential correlates of stimulus uncertainty. Science, 150, 1187–1188. Sutton, S., Tuetong, P., Zubin, J., & John, E. R. (1967). Information delivery and the sensory evoked potentials. Science, 155, 1436–1439. Van Petten, C., & Rheinfelder, H. 1995. Conceptual relationship between spoken words and environmental sounds: Event-related brain potential measures. Neuropsychologia, 33, 485–508. Watt, N. F. 1976. Two factor index of social position: Amherst modification. [An adaptation of Hollingshead & Redlich (1985)]. Unpublished manuscript, Amherst College, Department of Psychology. Wechsler, D. 1981. Wechsler Adult Intelligence Scale–Revised. San Antonio, TX: The Psychological Corp./Harcourt Brace Jovanovich. Woods, D. L., & Knight, R. T. 1986. Electrophysiological evidence of increased distractibility after dorsolateral prefrontal lesions. Neurology, 36, 212–216. Young, M. P., & Rugg, M. D. 1992. Word frequency and multiple repetition as determinants of the modulations of event-related potentials in a semantic classification task. Psychophysiology, 29, 664–676.