Perceptual and Motor Skills, 1976, 42, 655-661.
@ Percepmal and Motor Skills 1976
SIGNAL PULSE-RATEA N D JUDGED DURATION1 SANFORD GOLDSTONE A N D WILLIAM T. LHAMON Cofnell University Medical College Summary.-Theories which construct perception of time from content of input predict monotonic functions of rate-judged duration of stimuli, and do not account for intersensory differences. T w o experiments required Ss to compare directly the durations of paired lights o r sounds pulsed at various rates to produce discriminable beats and flickers (6.0, 10.0, 14.0 Hz), and steady signals. Pulsed lights, not sounds, were judged longer than steady, and this visual effect was identical for all flicker rates. Faster pulsed sounds and lights were judged longer than slower ones for all frequency combinations except for 10.0- to 14.0-Hz comparisons with vision. N o monotonic function of ratejudged durations of pulses was obtained; the effect was all-or-none. Although pulse rate did affect judged duration, neither simple functions nor symmetry across senses was found.
The fact that judgment of time has been related to the stimulus characteristics of the judged duration, e.g., sense mode (Goldstone, et d., 1959; Goldstone & Goldfarb, 1964), intensity of sound (Berglund, et al., 1969; Zelkind, 1973), and rate and number of input events (Fraisse, 1961; Frankenhaeuser, 1959; Matsuda & Matsuda, 1974), has led to models that construct temporal judgment directly from nontemporal factors: as an example, Ornstein (1969) builds temporal experience from the size of the information score required to contain events within an interval and suggests that judged duration is a monotonic function of storage size. Two experiments explored the plausibility of these simple models based upon magnitude of input by filling intervals with simple and easily discriminable pulsed auditory and visual signals of low frequencies to retain the perception of periodicity and higher frequencies for the perception of steadiness. Ss compared directly the duration of signals pulsed at different rates perceived as intermittent or steady. If judgment of time is constructed directly from amount of content of an interval (Frankenhaeuser, 1959; Ornstein, 1969), ( 1 ) magnitude of duration should be a monotonic function of the frequency of discriminable signal pulse races, ( 2 ) pulsed signals should be longer than steady for both audition and vision, and ( 3 ) these pulsed-steady differences should be a function of rate. If results are not compatible with these expectations, they argue for more complex temporal perceptual-cognitive systems (Goldstone, 1967; Treisman, 1963).
EXPERIMENT1 A recent study (Lhamon
& Goldstone, 1975) showed that flickering lights
lFrorn the Edward W. Bourne Behavioral Research Laboratories, 21 Bloomingdale Road, White Plains, New York 10605.
656
S. GOLDSTONE & W. T. LHAMON
(10 Hz) were judged longer than steady Iights (100 H z ) . This experiment employed two ocher easily discriminable frequencies, 6.0 and 14.0 Hz, and two sense modes, audition and vision, to determine: ( 1 ) if the flicker-steady difference is present with other rates, ( 2 ) if this difference is also present with audition, ( 3 ) if there is a monotonic relationship between signal pulse-rate and judged duration. Affirmative results would support theories of time judgment as solely constructed from event rate and storage size.
Method Using pair-comparison, Ss judged variable durations of flickering or steady lights, or, beating or steady sounds as shorter or longer than standard flickering or steady lights, or, beating or steady sounds. This yielded four stimulus-pair combinations for each sense mode. Visual-duration pairs consisted of flicker-flicker, steady-steady, flicker-steady. steady-flicker; auditory pairs were beat-beat, steady-steady, beat-steady, steady-beat. Ss judged the second duration of each pair as shorter or longer than the first using a 5-category response scale: 1, shorter; 2, slightly shorter; 3, equal; 4, slightly longer; 5, longer. Durations were on magnetic tape which controlled timing of auditory and visual signals, and intersc~rnuluand interpair intervals with accuracy and reliabiliry of better than 1.0% and 0 2 % respectively. The tape contained 105 pairs of durations distributed on two channels, one for auditory beat or visual flicker and the other for steady light or sound. The first member of each pair was 1.00 sec. followed by either 0.70, 0.85, 1.00, 1.15, or 1.30 sec. Five pairs were practice and 100 test pairs were organized haphazardly to permit five presenmrions of each of the five standard-variable combinations and the four stimulus sequences (vision: flicker-flicker, steady-steady, flicker-steady, steady-flicker; audition: beat-beat, steady-steady, beat-steady, steady-beat); interstimulus interval was 2.00 sec. and interpair interval 10.00 sec. Auditory durations were 1-KHz sine wave tones presenred dioticaily via headphones at 77 d b ( r e 2 X 10-6N/m2); the beats were created by repetitive pulses of sound at either 6.0 or 14.0 Hz having a 50% duty cycle such that a sound pulse was coincident with the beginning and end of all presenred durations. The visual source was an L.E.D. with peak emission at 655 nm (red) and luminance of 14 cd/m2 (77 d b ) ; Iights and sounds were equated for output (Stevens, 1955) and appeared subjectively equivalent in intensity with cross-modality matching. The steady light was pulsed at 100.0 Hz and the flickers at 6.0 and 14.0 Hz with viewing distance at 9 1 cm; target size was 4.6 mm or 0.29". Ss were studied in a sound-treated chamber with ambient sound pressure at 40 d b and luminance at 0.27 c d / m ' ( 5 9 d b ) . All lights and sounds were easily and comfortably perceived, and the two low frequencies gave clear flicker and bear; the signal pulse-rate difference between 6.0 and 14.0 Hz was easily detectable for both senses. Twenry volunteers age 19 to 53 yr. (Mdn 35 yr.) were divided into four equal groups: (1) audition, 6.0 Hz; ( 2 ) audition, 14.0 Hz; ( 3 ) vision, 6.0 Hz; ( 4 ) vision, 14.0 Hz. The Average Category Response (ACR) or mean response to each srandard-variable combination and each stimulus-pairing condition was calculated yielding the measure of magnitude of judged duration. The ACRs were examined with analysis of variance employing signal pulse-rate (6.0 or 14.0 Hz), sense mode (audition or vision), flicker-bear or steady judged, same (e.g., flicker-flicker, steady-steady) or different (e.g., flickersteady, steady-flicker) paired, and variable durations as the factors.
SIGNAL PULSE-RATE AND JUDGED DURATION
657
Results Fig. 1 displays the results of Exp. 1 for vision and audition combined over the two signal pulse rates, reflecting the interaction of sense mode X same or different paired X flicker-beat or steady judged X variable durations (F4/64 = 3.67, p < .01). The left panel shows that the pulsed lights were judged longer than steady lights, confirming Lharnon and Goldstone (1975) with two different rates; the right panel shows no difference between beating and steady sounds. Pulsed lights, not sounds, are judged longer than steady signals. The absence of any significant effects due to the signal pulse rate indicates no difference between judgments of 6.0- and 14.0-Hz durations. This experiment yields no simple relationship between number or rate of events and magnitude of judged duration. Durations of pulsed sounds were not judged longer than steady sounds of the same rate and equivalent intensity; although lights pulsed at the same rates as the sounds were judged longer than steady ones, this difference was no greater for the 14.0- than the 6.0-HZ flicker rate. r VISION
Variable durations (seconds1
FIG. 1. Average Category Response (ACR) X Variable Durations plots for audition and vision combined over the two pulse-rate groups, 6.0 and 14.0 Hz
EXPERIMENT2 This study explored more directly the effect of visual and auditory signalrate by requiring Ss to compare durations of light or sound pulsed at different rates. Three signal pulse-rates, 6.0, 10.0, and 14.0 Hz, were employed with pairs of durations of signals pulsed at two of the three frequencies; no steady lights or sounds were used. The direct comparison of two durations pulsed at different rates may represent a different temporal task than one employing steadypulsed comparisons, although predictions derived from storage size, and event rate or number schemata are the same; more events per unit time or a larger
S. GOLDSTONE & W. T.LHAMON
658
store should yield longer judged duration. This study bears directly upon models that yield predictions of judged duration as a monotonic function of event rate.
Method Using the same pair-comparison procedure, response scale, and duration control tape as in Exp. 1, Ss compared durations of lights o r sounds pulsed at one rate with durations pulsed at another rate. The durations of three rates were compared in pairs one with another in all combinations, and stimulus conditions and characteristics were identical with Exp. 1. Thirty volunteers age 1 8 to 56 yr. (Mdn 2 6 yr.) were divided into six equal groups, three audition and three vision, each having a common, recurrent signal pulse-rate (6.0, 10.0, 14.0 H z ) in each pair-comparison: ( 1 ) In the group where 6.0 Hz recurred in each of the five standard (1.00 sec.)-judged pairs of variable (0.70, 0.85, 1.00, 1.15, o r 1.30 sec.) durations, Ss compared durations pulsed at 6.0 with durations pulsed at 10.0 Hz, 10.00 with 6.0 Hz, 6.0 with 14.0 Hz, and 14.0 with 6.0 Hz; ( 2 ) where 10.0 H z recurred, Ss compared 10.0 with 6.0 Hz, 6.0 with 10.0 Hz, 10.0 with 14.0 Hz, and 14.0 with 10.0 Hz; ( 3 ) where 14.0 H z recurred, Ss compared 14.0 with 6.0 Hz, 6.0 with 14.0 Hz, 14.0 with 10.0 Hz, and 10.0 with 14.0 Hz. This design permitted the comparison of all combinations of the three signal pulse-rates and provided an internal replication for pulse-rate pairs since each pairing appeared separately in two of the three groups, e.g., 6.0- and 14.0-Hz comparisons were required in both the 6.0- and 14.0-Hz recurrent pulse-rate groups. Measures of Average Category Response were computed as in Exp. 1 and examined with analysis of variance: sense mode (audition o r vision), recurrent pulse-rate (6.0, 10.0, o r 14.0 H z ) , pulse-rate judged (higher of pair or lower of p a i r ) , pulse-rate of nonrecurrent frequency (higher of pair o r lower of pair), and variable durations were the factors.
Results Figs. 2 and 3 display the results for audition and vision respectively. Audition gave significant differences for all variable durations of all three pulse-rate 5.0 6.0- 14.OHz
10.0- 14.OM
6.0- 10.OHz
LL
3.0 P 01
??i
2.0
14.0- 10.OHz 14.0-6.OHz
a
10.0-6.OHz .70
1.00
1.30 .70 1.00 1.30 .70 Varlable durallons (seconds)
1.00
l.M
FIG. 2. Audition: Average Category Response ( A C R ) X Variable Durations plots for the three pulse-rate pairings combined over the recurrent frequency groups
SIGNAL PULSE-RATE AND JUDGED DURATION
6.0-10.OHz
6.0- 14.OHz
659
10.0- 14.OHz
14.0-6.OHz
4
14.0- 10.OHz
.70
1.W)
1.30 .70 1.00 1.30 .70 Variable durations (seconds)
1.00
1.30
FIG. 3. Vision: Average Category Response (ACR) X Variable Durations plots for the three pulse-race pairing combined over the recurrent frequency groups
pairs, 6.0-10.0 Hz, 6.0-14.0 Hz, and 10.0-14.0 Hz; vision showed the same for 6.0-10.0 Hz and 6.0-14.0 Hz; durations of higher signal pulse-rate were judged longer than those of lower rate (pulse-rate judged X recurrent pulserate X pulse-rate of nonrecurrent frequency, F2/.'4 = 8.13, p < .001) except for the 10.0- to 14.0-Hz visual comparisons (sense mode X pulse-rate judged X pulse-rate of nonrecurrent frequency, F l I z 4 = 13.50, p < .001) where no signal pulse-rate effect was obtained. Each of the three pulse-rate pairing combinations was received by 10 Ss with internal replication provided via presence in two equally divided recurrent pulse-rate groups. Nine of 10 auditory and 10 of 10 visual Ss yielded longer Average Category Responses when comparing durations of 14.0 with 6.0 Hz than 6.0 with 14.0 Hz; nine of 10 auditory and 10 of 10 visual Ss yielded longer Average Category Responses for 10.0 compared with 6.0 Hz than 6.0 compared with 10.0 Hz; 9 of 10 auditory Ss yielded longer Average Category Responses for 14.0 compared with 10.0 Hz than 10.0 with 14.0 Hz, but there was no trend for vision. The presence of an auditory signal pulse-rate effect on judged duration here but not in Exp. 1 suggests that comparison of pulsed sounds of different rates involves different temporal processing from comparison of pulsed with steady sounds; the absence of a difference in judged duration between two easily discriminable visual pulse-rates, 10.0 and 14.0 Hz, suggests a limit upon the pulse rate-judged duration relationship; the absence of a pulse-rate d i f f e r e n c e judged-duration function reduces the likelihood that judged duration is a monotonic function of number or rate of stimulus events.
DISCUSSION These direct-comparison experiments showed again the importance of stimulus rate in the judgment of duration. However, this effect is complex
660
S. GOLDSTONE & W. T. LHAMON
and it is unlikely that it can be explained by a simple function of rate-judged duration for input. Two distinct rates of pulsed visual signals were judged longer than steady lights. If judged duration is constructed from rate and amount of input, the same effect should emerge with auditory signals of equivalent intensity and pulsed at the same rate; pulsed sounds were not judged longer than steady. If judged duration is derived functionally from stimulus frequency and number, the faster visual rate should have differed most from the steady lights, but no difference berween faster and slower rates was obtained. Similarly, direct comparisons of duration among three distinct auditory and visual frequencies yielded no functional relationship between the amount of pulserate difference in a pair and judged duration. Finally, since information processing and memory schemata (Frankenhaeuser, 1959; Ornstein, 1969) provide no plausible basis for intersensory and intrasensory subtleties, they cannot explain the facts that: ( 1 ) two rates of pulsed sounds were not different from steady auditory signals, while the same faster pulsed sounds were judged longer than slower ones; ( 2 ) direct comparison of two signal pulse-rates, 10.0 and 14.0 Hz, produced an effect for audition but none was shown with vision. Generalizations from these experiments must be limited to this range of signal pulse-rates all of which were easily discriminable one from the other, this narrow range of short durations, and this psychophysical method. The same should be required of other investigations from which theories are constructed. Studies of human judgment of time have shown biological effects (e.g., Goldstone & Kirkham, 1968; Hoagland, 1933; Kleber, et &., 1963) compatible with an "internal clock" hypothesis; other research has shown the need for concern about intersensory differences (e.g., Goldstone, et al., 1959; Goldstone & Lhamon, 1971) and stimulus effects (e.g., Berglund, et al., 1969; Zelkind, 1973); still others emphasize learned, cognitive factors (e.g., Frankenhaeuser, 1959; Ornstein, 1969). It is our view that the evidence from prior work and these experiments support a model of human temporal judgment and experience that incorporates biologic, i.e., "internal clock," sensory-perceptual, and cognitive factors in a single schema (e.g., Treisman, 1963); emphasis upon one or another factor is determined by the specific issue for investigation. The judgment of time by humans is seen as a complex, evolving natural function, the study and understanding of which requires the use of multiple frames of reference. REFERENCES BERGLUND,B., BERGLUND, U., E ~ A NG., , & FRANKENHABUSER,M. The influence of auditory stimulus intensity on apparent duration. Scand. J. Psychol., 1969, 10, 2 1-26.
Influence de la durie et de la frequence des changements sur I'estimation du temps. LJAnnBe Psychol., 1961, 61, 325-340. FRANKENHAEUSER, M. Estimation of time: an experimental study. Stockholm: Almqvist & Wiksell, 1959.
FRAISSE, P.
SIGNAL PULSE-RATE AND JUDGED DURATION
66 1
GOLDSTONE,S. The human clock: a framework for the study of healthy and deviant time erception. In R. Fischer (Ed.), Interdisciplinary perspectives of time. A nna?s N.Y. Acad. Sci., 1967, 138, 767-783. GOLDSTONE,S., BOARDMAN,W. K, & LHAMON, W. T. Intersensory comparisons of temporal judgments. I . exp. Psychol., 1959, 57, 243-248. GOLDSTONE, S., & GOLDFARB,J. L. Direct comparison of auditory and visual durations. I . exp. Psychol., 1964, 67, 483-485. GOLDSTONE, S., & KIRKHMI, J. E. The effects of secobarbital and dextroamphetamine upon time judgment: intersensory factors. Psychopharmacologia, 1968, 13, 65-73. GOLDSTONE,S., & LHAMON,W. T. Levels of cognirive functioning and the auditoryvisual differences in human timing behavior. In M. Appley (Ed.), Adaptationlevel lheory (symposium). New York: Academic Press, 1971. Pp. 263-280. HOAGLAND,H. The physiological control of judgments of duration: evidence for a chemical clock. I . gen. Psychol., 1933, 9, 267-287. KLEBER, R. J., LHAMON,W. T., & GOLDSTONE, S. Hyperthermia, hyperthyroidism, and time judgment. I . cornp. phyriol. Psychol., 1963, 56, 362-365. LHAMON,W. T., & GOLDSTONE, S. Movement and the judged duration of visual targets. Bull. Psychon. Soc., 1975, 5, 53-54. MATSUDA,F., & MATSUDA,M. Effects of frequency of intermittent stimuli on time estimation in children and in adults: I. Sounds and lights. Psychologia, 1974, 17, 206-212. ORNSTEIN,R. E. O n the experience o f time. Middlesex: Penguin Books, 1969. TREISMAN, M. Temporal discrimination and the indifference interval: implications for a model of the "internal clock." Psychol. Monogs., 1963, 77, No. 13 (Whole No. 576). ZELKIND, I. Factors in time estimation and a case for the internal clock. I . gen. Psychol., 1973,88, 295-301. Accepted January 12, 1976.
@ Percepmal and Motor Skills 1976
SIGNAL PULSE-RATEA N D JUDGED DURATION1 SANFORD GOLDSTONE A N D WILLIAM T. LHAMON Cofnell University Medical College Summary.-Theories which construct perception of time from content of input predict monotonic functions of rate-judged duration of stimuli, and do not account for intersensory differences. T w o experiments required Ss to compare directly the durations of paired lights o r sounds pulsed at various rates to produce discriminable beats and flickers (6.0, 10.0, 14.0 Hz), and steady signals. Pulsed lights, not sounds, were judged longer than steady, and this visual effect was identical for all flicker rates. Faster pulsed sounds and lights were judged longer than slower ones for all frequency combinations except for 10.0- to 14.0-Hz comparisons with vision. N o monotonic function of ratejudged durations of pulses was obtained; the effect was all-or-none. Although pulse rate did affect judged duration, neither simple functions nor symmetry across senses was found.
The fact that judgment of time has been related to the stimulus characteristics of the judged duration, e.g., sense mode (Goldstone, et d., 1959; Goldstone & Goldfarb, 1964), intensity of sound (Berglund, et al., 1969; Zelkind, 1973), and rate and number of input events (Fraisse, 1961; Frankenhaeuser, 1959; Matsuda & Matsuda, 1974), has led to models that construct temporal judgment directly from nontemporal factors: as an example, Ornstein (1969) builds temporal experience from the size of the information score required to contain events within an interval and suggests that judged duration is a monotonic function of storage size. Two experiments explored the plausibility of these simple models based upon magnitude of input by filling intervals with simple and easily discriminable pulsed auditory and visual signals of low frequencies to retain the perception of periodicity and higher frequencies for the perception of steadiness. Ss compared directly the duration of signals pulsed at different rates perceived as intermittent or steady. If judgment of time is constructed directly from amount of content of an interval (Frankenhaeuser, 1959; Ornstein, 1969), ( 1 ) magnitude of duration should be a monotonic function of the frequency of discriminable signal pulse races, ( 2 ) pulsed signals should be longer than steady for both audition and vision, and ( 3 ) these pulsed-steady differences should be a function of rate. If results are not compatible with these expectations, they argue for more complex temporal perceptual-cognitive systems (Goldstone, 1967; Treisman, 1963).
EXPERIMENT1 A recent study (Lhamon
& Goldstone, 1975) showed that flickering lights
lFrorn the Edward W. Bourne Behavioral Research Laboratories, 21 Bloomingdale Road, White Plains, New York 10605.
656
S. GOLDSTONE & W. T. LHAMON
(10 Hz) were judged longer than steady Iights (100 H z ) . This experiment employed two ocher easily discriminable frequencies, 6.0 and 14.0 Hz, and two sense modes, audition and vision, to determine: ( 1 ) if the flicker-steady difference is present with other rates, ( 2 ) if this difference is also present with audition, ( 3 ) if there is a monotonic relationship between signal pulse-rate and judged duration. Affirmative results would support theories of time judgment as solely constructed from event rate and storage size.
Method Using pair-comparison, Ss judged variable durations of flickering or steady lights, or, beating or steady sounds as shorter or longer than standard flickering or steady lights, or, beating or steady sounds. This yielded four stimulus-pair combinations for each sense mode. Visual-duration pairs consisted of flicker-flicker, steady-steady, flicker-steady. steady-flicker; auditory pairs were beat-beat, steady-steady, beat-steady, steady-beat. Ss judged the second duration of each pair as shorter or longer than the first using a 5-category response scale: 1, shorter; 2, slightly shorter; 3, equal; 4, slightly longer; 5, longer. Durations were on magnetic tape which controlled timing of auditory and visual signals, and intersc~rnuluand interpair intervals with accuracy and reliabiliry of better than 1.0% and 0 2 % respectively. The tape contained 105 pairs of durations distributed on two channels, one for auditory beat or visual flicker and the other for steady light or sound. The first member of each pair was 1.00 sec. followed by either 0.70, 0.85, 1.00, 1.15, or 1.30 sec. Five pairs were practice and 100 test pairs were organized haphazardly to permit five presenmrions of each of the five standard-variable combinations and the four stimulus sequences (vision: flicker-flicker, steady-steady, flicker-steady, steady-flicker; audition: beat-beat, steady-steady, beat-steady, steady-beat); interstimulus interval was 2.00 sec. and interpair interval 10.00 sec. Auditory durations were 1-KHz sine wave tones presenred dioticaily via headphones at 77 d b ( r e 2 X 10-6N/m2); the beats were created by repetitive pulses of sound at either 6.0 or 14.0 Hz having a 50% duty cycle such that a sound pulse was coincident with the beginning and end of all presenred durations. The visual source was an L.E.D. with peak emission at 655 nm (red) and luminance of 14 cd/m2 (77 d b ) ; Iights and sounds were equated for output (Stevens, 1955) and appeared subjectively equivalent in intensity with cross-modality matching. The steady light was pulsed at 100.0 Hz and the flickers at 6.0 and 14.0 Hz with viewing distance at 9 1 cm; target size was 4.6 mm or 0.29". Ss were studied in a sound-treated chamber with ambient sound pressure at 40 d b and luminance at 0.27 c d / m ' ( 5 9 d b ) . All lights and sounds were easily and comfortably perceived, and the two low frequencies gave clear flicker and bear; the signal pulse-rate difference between 6.0 and 14.0 Hz was easily detectable for both senses. Twenry volunteers age 19 to 53 yr. (Mdn 35 yr.) were divided into four equal groups: (1) audition, 6.0 Hz; ( 2 ) audition, 14.0 Hz; ( 3 ) vision, 6.0 Hz; ( 4 ) vision, 14.0 Hz. The Average Category Response (ACR) or mean response to each srandard-variable combination and each stimulus-pairing condition was calculated yielding the measure of magnitude of judged duration. The ACRs were examined with analysis of variance employing signal pulse-rate (6.0 or 14.0 Hz), sense mode (audition or vision), flicker-bear or steady judged, same (e.g., flicker-flicker, steady-steady) or different (e.g., flickersteady, steady-flicker) paired, and variable durations as the factors.
SIGNAL PULSE-RATE AND JUDGED DURATION
657
Results Fig. 1 displays the results of Exp. 1 for vision and audition combined over the two signal pulse rates, reflecting the interaction of sense mode X same or different paired X flicker-beat or steady judged X variable durations (F4/64 = 3.67, p < .01). The left panel shows that the pulsed lights were judged longer than steady lights, confirming Lharnon and Goldstone (1975) with two different rates; the right panel shows no difference between beating and steady sounds. Pulsed lights, not sounds, are judged longer than steady signals. The absence of any significant effects due to the signal pulse rate indicates no difference between judgments of 6.0- and 14.0-Hz durations. This experiment yields no simple relationship between number or rate of events and magnitude of judged duration. Durations of pulsed sounds were not judged longer than steady sounds of the same rate and equivalent intensity; although lights pulsed at the same rates as the sounds were judged longer than steady ones, this difference was no greater for the 14.0- than the 6.0-HZ flicker rate. r VISION
Variable durations (seconds1
FIG. 1. Average Category Response (ACR) X Variable Durations plots for audition and vision combined over the two pulse-rate groups, 6.0 and 14.0 Hz
EXPERIMENT2 This study explored more directly the effect of visual and auditory signalrate by requiring Ss to compare durations of light or sound pulsed at different rates. Three signal pulse-rates, 6.0, 10.0, and 14.0 Hz, were employed with pairs of durations of signals pulsed at two of the three frequencies; no steady lights or sounds were used. The direct comparison of two durations pulsed at different rates may represent a different temporal task than one employing steadypulsed comparisons, although predictions derived from storage size, and event rate or number schemata are the same; more events per unit time or a larger
S. GOLDSTONE & W. T.LHAMON
658
store should yield longer judged duration. This study bears directly upon models that yield predictions of judged duration as a monotonic function of event rate.
Method Using the same pair-comparison procedure, response scale, and duration control tape as in Exp. 1, Ss compared durations of lights o r sounds pulsed at one rate with durations pulsed at another rate. The durations of three rates were compared in pairs one with another in all combinations, and stimulus conditions and characteristics were identical with Exp. 1. Thirty volunteers age 1 8 to 56 yr. (Mdn 2 6 yr.) were divided into six equal groups, three audition and three vision, each having a common, recurrent signal pulse-rate (6.0, 10.0, 14.0 H z ) in each pair-comparison: ( 1 ) In the group where 6.0 Hz recurred in each of the five standard (1.00 sec.)-judged pairs of variable (0.70, 0.85, 1.00, 1.15, o r 1.30 sec.) durations, Ss compared durations pulsed at 6.0 with durations pulsed at 10.0 Hz, 10.00 with 6.0 Hz, 6.0 with 14.0 Hz, and 14.0 with 6.0 Hz; ( 2 ) where 10.0 H z recurred, Ss compared 10.0 with 6.0 Hz, 6.0 with 10.0 Hz, 10.0 with 14.0 Hz, and 14.0 with 10.0 Hz; ( 3 ) where 14.0 H z recurred, Ss compared 14.0 with 6.0 Hz, 6.0 with 14.0 Hz, 14.0 with 10.0 Hz, and 10.0 with 14.0 Hz. This design permitted the comparison of all combinations of the three signal pulse-rates and provided an internal replication for pulse-rate pairs since each pairing appeared separately in two of the three groups, e.g., 6.0- and 14.0-Hz comparisons were required in both the 6.0- and 14.0-Hz recurrent pulse-rate groups. Measures of Average Category Response were computed as in Exp. 1 and examined with analysis of variance: sense mode (audition o r vision), recurrent pulse-rate (6.0, 10.0, o r 14.0 H z ) , pulse-rate judged (higher of pair or lower of p a i r ) , pulse-rate of nonrecurrent frequency (higher of pair o r lower of pair), and variable durations were the factors.
Results Figs. 2 and 3 display the results for audition and vision respectively. Audition gave significant differences for all variable durations of all three pulse-rate 5.0 6.0- 14.OHz
10.0- 14.OM
6.0- 10.OHz
LL
3.0 P 01
??i
2.0
14.0- 10.OHz 14.0-6.OHz
a
10.0-6.OHz .70
1.00
1.30 .70 1.00 1.30 .70 Varlable durallons (seconds)
1.00
l.M
FIG. 2. Audition: Average Category Response ( A C R ) X Variable Durations plots for the three pulse-rate pairings combined over the recurrent frequency groups
SIGNAL PULSE-RATE AND JUDGED DURATION
6.0-10.OHz
6.0- 14.OHz
659
10.0- 14.OHz
14.0-6.OHz
4
14.0- 10.OHz
.70
1.W)
1.30 .70 1.00 1.30 .70 Variable durations (seconds)
1.00
1.30
FIG. 3. Vision: Average Category Response (ACR) X Variable Durations plots for the three pulse-race pairing combined over the recurrent frequency groups
pairs, 6.0-10.0 Hz, 6.0-14.0 Hz, and 10.0-14.0 Hz; vision showed the same for 6.0-10.0 Hz and 6.0-14.0 Hz; durations of higher signal pulse-rate were judged longer than those of lower rate (pulse-rate judged X recurrent pulserate X pulse-rate of nonrecurrent frequency, F2/.'4 = 8.13, p < .001) except for the 10.0- to 14.0-Hz visual comparisons (sense mode X pulse-rate judged X pulse-rate of nonrecurrent frequency, F l I z 4 = 13.50, p < .001) where no signal pulse-rate effect was obtained. Each of the three pulse-rate pairing combinations was received by 10 Ss with internal replication provided via presence in two equally divided recurrent pulse-rate groups. Nine of 10 auditory and 10 of 10 visual Ss yielded longer Average Category Responses when comparing durations of 14.0 with 6.0 Hz than 6.0 with 14.0 Hz; nine of 10 auditory and 10 of 10 visual Ss yielded longer Average Category Responses for 10.0 compared with 6.0 Hz than 6.0 compared with 10.0 Hz; 9 of 10 auditory Ss yielded longer Average Category Responses for 14.0 compared with 10.0 Hz than 10.0 with 14.0 Hz, but there was no trend for vision. The presence of an auditory signal pulse-rate effect on judged duration here but not in Exp. 1 suggests that comparison of pulsed sounds of different rates involves different temporal processing from comparison of pulsed with steady sounds; the absence of a difference in judged duration between two easily discriminable visual pulse-rates, 10.0 and 14.0 Hz, suggests a limit upon the pulse rate-judged duration relationship; the absence of a pulse-rate d i f f e r e n c e judged-duration function reduces the likelihood that judged duration is a monotonic function of number or rate of stimulus events.
DISCUSSION These direct-comparison experiments showed again the importance of stimulus rate in the judgment of duration. However, this effect is complex
660
S. GOLDSTONE & W. T. LHAMON
and it is unlikely that it can be explained by a simple function of rate-judged duration for input. Two distinct rates of pulsed visual signals were judged longer than steady lights. If judged duration is constructed from rate and amount of input, the same effect should emerge with auditory signals of equivalent intensity and pulsed at the same rate; pulsed sounds were not judged longer than steady. If judged duration is derived functionally from stimulus frequency and number, the faster visual rate should have differed most from the steady lights, but no difference berween faster and slower rates was obtained. Similarly, direct comparisons of duration among three distinct auditory and visual frequencies yielded no functional relationship between the amount of pulserate difference in a pair and judged duration. Finally, since information processing and memory schemata (Frankenhaeuser, 1959; Ornstein, 1969) provide no plausible basis for intersensory and intrasensory subtleties, they cannot explain the facts that: ( 1 ) two rates of pulsed sounds were not different from steady auditory signals, while the same faster pulsed sounds were judged longer than slower ones; ( 2 ) direct comparison of two signal pulse-rates, 10.0 and 14.0 Hz, produced an effect for audition but none was shown with vision. Generalizations from these experiments must be limited to this range of signal pulse-rates all of which were easily discriminable one from the other, this narrow range of short durations, and this psychophysical method. The same should be required of other investigations from which theories are constructed. Studies of human judgment of time have shown biological effects (e.g., Goldstone & Kirkham, 1968; Hoagland, 1933; Kleber, et &., 1963) compatible with an "internal clock" hypothesis; other research has shown the need for concern about intersensory differences (e.g., Goldstone, et al., 1959; Goldstone & Lhamon, 1971) and stimulus effects (e.g., Berglund, et al., 1969; Zelkind, 1973); still others emphasize learned, cognitive factors (e.g., Frankenhaeuser, 1959; Ornstein, 1969). It is our view that the evidence from prior work and these experiments support a model of human temporal judgment and experience that incorporates biologic, i.e., "internal clock," sensory-perceptual, and cognitive factors in a single schema (e.g., Treisman, 1963); emphasis upon one or another factor is determined by the specific issue for investigation. The judgment of time by humans is seen as a complex, evolving natural function, the study and understanding of which requires the use of multiple frames of reference. REFERENCES BERGLUND,B., BERGLUND, U., E ~ A NG., , & FRANKENHABUSER,M. The influence of auditory stimulus intensity on apparent duration. Scand. J. Psychol., 1969, 10, 2 1-26.
Influence de la durie et de la frequence des changements sur I'estimation du temps. LJAnnBe Psychol., 1961, 61, 325-340. FRANKENHAEUSER, M. Estimation of time: an experimental study. Stockholm: Almqvist & Wiksell, 1959.
FRAISSE, P.
SIGNAL PULSE-RATE AND JUDGED DURATION
66 1
GOLDSTONE,S. The human clock: a framework for the study of healthy and deviant time erception. In R. Fischer (Ed.), Interdisciplinary perspectives of time. A nna?s N.Y. Acad. Sci., 1967, 138, 767-783. GOLDSTONE,S., BOARDMAN,W. K, & LHAMON, W. T. Intersensory comparisons of temporal judgments. I . exp. Psychol., 1959, 57, 243-248. GOLDSTONE, S., & GOLDFARB,J. L. Direct comparison of auditory and visual durations. I . exp. Psychol., 1964, 67, 483-485. GOLDSTONE, S., & KIRKHMI, J. E. The effects of secobarbital and dextroamphetamine upon time judgment: intersensory factors. Psychopharmacologia, 1968, 13, 65-73. GOLDSTONE,S., & LHAMON,W. T. Levels of cognirive functioning and the auditoryvisual differences in human timing behavior. In M. Appley (Ed.), Adaptationlevel lheory (symposium). New York: Academic Press, 1971. Pp. 263-280. HOAGLAND,H. The physiological control of judgments of duration: evidence for a chemical clock. I . gen. Psychol., 1933, 9, 267-287. KLEBER, R. J., LHAMON,W. T., & GOLDSTONE, S. Hyperthermia, hyperthyroidism, and time judgment. I . cornp. phyriol. Psychol., 1963, 56, 362-365. LHAMON,W. T., & GOLDSTONE, S. Movement and the judged duration of visual targets. Bull. Psychon. Soc., 1975, 5, 53-54. MATSUDA,F., & MATSUDA,M. Effects of frequency of intermittent stimuli on time estimation in children and in adults: I. Sounds and lights. Psychologia, 1974, 17, 206-212. ORNSTEIN,R. E. O n the experience o f time. Middlesex: Penguin Books, 1969. TREISMAN, M. Temporal discrimination and the indifference interval: implications for a model of the "internal clock." Psychol. Monogs., 1963, 77, No. 13 (Whole No. 576). ZELKIND, I. Factors in time estimation and a case for the internal clock. I . gen. Psychol., 1973,88, 295-301. Accepted January 12, 1976.
