the tritone paradox: a link between music and speech

8/2/2019 The Tritone Paradox: A Link between Music and Speech

1/8

The Tritone Paradox: A Link between Music and SpeechAuthor(s): Diana DeutschReviewed work(s):Source: Current Directions in Psychological Science, Vol. 6, No. 6 (Dec., 1997), pp. 174-180Published by: Sage Publications, Inc. on behalf of Association for Psychological ScienceStable URL: http://www.jstor.org/stable/20182482 .

Accessed: 16/03/2012 10:08

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of

content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms

of scholarship. For more information about JSTOR, please contact [email protected].

Sage Publications, Inc. andAssociation for Psychological Science are collaborating with JSTOR to digitize,

preserve and extend access to Current Directions in Psychological Science.

htt // j t
http://www.jstor.org/action/showPublisher?publisherCode=sagehttp://www.jstor.org/action/showPublisher?publisherCode=assocpsychscihttp://www.jstor.org/stable/20182482?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/20182482?origin=JSTOR-pdfhttp://www.jstor.org/action/showPublisher?publisherCode=assocpsychscihttp://www.jstor.org/action/showPublisher?publisherCode=sage


2/8

174 VOLUME 6,NUMBER 6, DECEMBER 1997

Pasek, K. (1996). Infants' sensitivity to wordboundaries in fluent speech. Journal of ChildLanguage, 23, 1-30.Newsome, M., & Jusczyk, PW. (1995). Do infantsuse stress as a cue for segmenting fluentspeech? In D. MacLaughlin & S. McEwen(Eds.), Proceedings of the 19th Annual Boston

University Conference on Language Development(Vol. 2, pp. 415^26). Somerville, MA: Cascadilla Press.Saffran, J.R., Aslin, R.N., & Newport, E.L. (1996).Statistical learning by 8-month-old infants. Science, 274, 1926-1928.Suomi, K. (1993). An outline of a developmental

model of adult phonological organization andbehavior. Journal of Phonetics, 21, 29-60.Woodward, J.Z., & Aslin, R.N. (1990, April). Segmentation cues inmaternal speech to infants. Paper presented at the biennial meeting of theInternational Conference on Infant Studies,

Montreal, Quebec.

The Tritone Paradox: A Link BetweenMusic and Speech

Diana Deutsch1Department of Psychology, University of California, San Diego, La Jolla, California

When William Harvey, in his Iclassic treatise on the circulation ofthe blood, declared that pulsesounds could be heard in the chest,this assertion was hotly disputed.

During the controversy that followed, a Venetian physician jokingly declared that perhaps suchsounds were "only to be heard inLondon" (Hunt, 1978, p. 138).

The idea that certain soundsmight be perceived quite differently in different geographic regions appears so unlikely as toform the basis of a joke. Yet thisarticle concerns a sound patternthat, it strongly appears, is indeedheard differently by people in different geographic regions. I explorethe properties of this pattern andthe basis for this curious geographic association.

PITCH STRUCTURES

Pitch?the attribute of "highness" or "lowness" of sound?iscentral to our perception of music.

Although many pitches that we encounter outside music are ratherblurred (consider, e.g., footsteps,squeaks, and rustling sounds),those produced by most musicalinstruments and the singing voiceare clearly defined. It is from suchpitches that melodies and harmonies are formed.

Questions concerning pitch andpitch relationships have intriguedscientists since ancient times.Pythagoras is credited with showing experimentally that the pitch ofa vibrating string varies inverselywith its length, and that particularmusical intervals are producedwhen string lengths stand in certain ratios. During the 17th century, Galileo and Mersenneshowed that these associations arebased on the relationship betweenstring length and vibration frequency: The shorter the string, thehigher the rate of vibration.Mersenne also discovered thatwhen a body vibrates, it does sonot only at the frequency that corresponds to its perceived pitch?

the fundamental frequency?butalso at frequencies that are integermultiples of the fundamental. Sucha set of frequencies is called a harmonic series. Later Seebeck (1843)and Schouten (1940) showed that

when presented with a harmonicseries, the listener hears a pitch thatcorresponds to the fundamental,even when the fundamental itselfisweak or absent.

How do we abstract relationships between pitches so as to perceive musical patterns? When twotones are presented in combination, we perceive a musical interval, and we perceive intervals asbeing the same in size when theircomponent frequencies stand inthe same ratio. The traditional musical scale is based in part on thisprinciple. Two adjacent notes on akeyboard form a pitch relationshipcalled a semitone; this correspondsto a frequency ratio of approxi

mately 18:17. Intervals that comprise the same number of semitones are given the same name. Forexample, an interval comprising 12semitones (a ratio of 2:1) is calledan octave, an interval comprising 7semitones (a ratio of 3:2) is called aperfect fifth, and an interval comprising 5 semitones (a ratio of 4:3)is called a perfect fourth.Because of the perceptualequivalence of tone pairs that formthe same interval, a musical passage can be played by means of different pitches (i.e., it can be transposed to different keys), andprovided that the intervals remainthe same, its perceptual identity ispreserved. Musical patterns are inthis respect like visual shapes,

Recommended ReadingDeutsch, D. (1991). (See References)Deutsch, D. (1996). The perception of

auditory patterns. InW. Prinz &B. Bridgeman (Eds.)/ Handbook ofperception and action (Vol. 1, pp.253-296). New York: Wiley.Deutsch, D. (Ed.). (in press). The psychology of music (2nd ed.). San Diego: Academic Press.

Published by Cambridge University Press


3/8

CURRENT DIRECTIONS IN PSYCHOLOGICALSCIENCE 175

which retain their perceptual identities when they are translated todifferent positions in the visualfield. Indeed, the notion that such apattern might

beperceived

as radically different under transpositionis as paradoxical as the notion thata visual shape might undergo a

metamorphosis through beingmoved to a different position inspace.

Other pattern analyses are basedon the octave. It has long been recognized that tones that are relatedby octaves have a certain perceptual equivalence. This is acknowledged in the system of notation forthe traditional musical scale. Thecore of this scale consists of 12tones, corresponding to the division of the octave into semitones,and each tone is given a name (C,C#, D, and so on). The entire scale,as it ascends in height (as in goingfrom left to right up a keyboard),consists of the repeating occurrence of this sequence of notenames across successive octaves.Octave placement is designated bysubscripts, and tones that stand inoctave relation are held to be in thesame pitch class. For example,tones C3, C4, and C5 are all in pitchclass C, but in successively higheroctaves; tones D#3, D#4, and D#5are all in pitch class D#, but in successively higher octaves; and so on.We can thus regard the pitch ofa tone as varying along two dimensions: The monotonie dimension ofheight defines its position on a continuum from low to high, and thecircular dimension of pitch classdefines its position within the octave. We may then ask whetherthese two dimensions are independent, or whether they interact insome fashion. Common sensewould say that surely they must be

independent. If a musician wereasked, "Which tone is higher, C orF#?" a likely reply is that the question is meaningless: One wouldneed to know which C and which F#

before a reasonable answer couldbe given.

ENDLESSLY RISING ANDFALLING PITCHES

Shepard (1964) performed an experiment that he hypothesizedwould demonstrate the independence of the dimensions of pitchclass and pitch height. He generated a series of tones, each of whichconsisted of a set of pure-tone components that were separated by octaves.2 For example, one tone consisted of components ..., C2, C3,

C4, . . ., and so on, and anotherconsisted of components . .., C#2,

C#3, C#4, . . ., and so on. In thisway, the tones were clearly definedin terms of pitch class, but their octave placement was ambiguous.For each tone, the amplitudes ofthe components were determinedby a bell-shaped spectral envelope,so that those in the middle of the

musical range were highest, andthose at the extremes were lowest.Shepard varied the pitch classes ofthe tones while keeping the position and shape of the envelope constant.

Shepard found that when twosuch tones were played in succession, listeners heard either an ascending pattern or a descendingone, depending on which was theshorter distance along the pitchclass circle (Fig. 1). So, for example,the pair D#-E was always heard asascending, because the shorter distance was clockwise. Similarly, the

pair A#-A was always heard as descending, because the shorter distance was counterclockwise. Shepard then produced a fascinatingdemonstration: A series of tonesthat repeatedly moved all aroundthe pitch class circle in clockwisedirection appeared to be endlesslyascending. When, alternatively, thetones moved around the circle in

rv^v_^x^c#)(a#) (jT)vD (3)

Fig. 1. The pitch class circle, formed bythe 12 pitch classes within the octave.To produce the tritone paradox, tonesthat are in opposite positions along thecircle are played sequentially; for example, D followed by G# might beplayed, or F followed by B.

counterclockwise steps, the seriesappeared to descend endlessly.This musical paradox has a visualcounterpart in the endless staircasedevised by Penrose and Penrose(1958), and later popularized bythe artist M.C. Escher.

The composer Jean-Claude Risset produced a number of compelling variants of this pitch paradox(Risset, 1971). In one case, a singletone glided around the pitch classcircle in clockwise direction, so thatit seemed to ascend endlessly.Counterclockwise movement produced the impression of an endlessly descending glide. In another

demonstration, a tone glided clockwise around the pitch class circlewhile its spectral envelope moveddownward, so that the tone appeared both to ascend and to descend at the same time.

At first sight, such demonstrations of pitch circularity appear toconfirm the hypothesis that pitchclass and pitch height are independent. However, another influenceis operating in this situation also.In grouping together elements of a

perceptual array, we tend to formlinkages between those that areclose together, in preference tothose that are further apart. Thereare several well-known illustra

Copyright ? 1998 American Psychological Society


4/8

176 VOLUME 6,NUMBER 6, DECEMBER 1997

tions of this principle in vision,such as the tendency to group together dots that are positioned inspatial proximity, or to perceivemotion between adjacent lightsthat come on and off in succession(Rock, 1986). In the case of music,we tend to create perceptual linkages between tones that form smallmelodic intervals rather than largeones (Bregman & Campbell, 1971;

Deutsch, 1975; Van Noorden,1975). This raises the possibilitythat in pitch circularity demonstrations, judgments of relative height

might be so heavily influenced byproximity that a potential association between pitch class and perceived height would be overwhelmed by this factor.

THE TRITONE PARADOX

Given such reasoning, I wondered how listeners would judge apair of such tones that were relatedby exactly a half-octave (this interval is called a tritone), so that theywere separated by the same distance along the pitch class circle ineither direction. What would happen, for example, ifD were played,followed by G#, or if B were followed by F? Because proximitycould not then be invoked by theperceptual system, would judgments of relative height be ambiguous, as predicted from the assumption of independence, or would an

association between pitch class andperceived height emerge? It occurred to me that in making suchjudgments, the listener could referto the absolute positions of thetones along the pitch class circle. Iftones in one region of the circlewere tagged as higher, and those inthe opposite region were tagged aslower, the ambiguity of heightwould be resolved. Imagine, for example, that the listener representedthe pitch class circle as shown inFigure 1,with the vertical axis representing perceived height. If thelistener mentally placed pitch classC in the highest position (as in thisfigure), then he or she would hearC followed by F# as descending. If,however, the listener mentallyplaced F# in the highest positioninstead, then he or she would hearthe identical pair of tones as ascending.This reasoning led me to embarkon a series of experiments that employed the following procedure.Subjects were presented with tritone pairs such as Ihave described,and they judged whether each pairformed an ascending or a descending pattern.3 The percentage oftimes that the subject heard a pattern as descending was then plotted as a function of the pitchclass of the first tone of the pair(Deutsch, 1986, 1987, 1991, 1994,1996; Deutsch, Kuyper, & Fisher,1987; Deutsch, North, & Ray, 1990;

Ragozzine & Deutsch, 1994).Each tone consisted of six puretone components that were separated by octaves, and whose amplitudes were determined by abell-shaped spectral envelope. (Aspectral representation of such atone pair is shown in Fig. 2.) Thetone pairs were generated under,envelopes that were placed at different positions along the spectrum, in most experiments spacedat half-octave intervals. Varyingthe envelope positions in this fashion controlled for interpretations

_I _I_

-201/ \ I

-40-1 / I \ I: "\ M I 1 I 1 1 \i i i *? jLU* i-'-'-n

LUy/\ ^v

i1 / \

I-A_11 1 1 1 1 11 \i?? ?*-?i100 1000FREQUENCY (Hz)

Fig. 2. Spectral representation of a tone pair that produces the tritone paradox. Theupper graph represents a tone of pitch class D, and the lower graph represents a toneof pitch class G#. I



5/8

CURRENT DIRECTIONS IN PSYCHOLOGICALSCIENCE 177

based on the relative amplitudes orloudnesses of the harmonic components of the tones.Figure 3 shows the judgments of4 different subjects using this para

digm. In each case, the data wereaveraged over four spectral envelopes that were spaced at halfoctave intervals. It can be seen thatthe judgments of each subject depended systematically on the positions of the tones along the pitchclass circle. However, the directionof this dependence varied strikingly from one subject to another.So we can think of the pitch classcircle as having a particular orientation with respect to height, andthis orientation differs from one

person to another. For example, thesubject whose data are shown onthe upper right of Figure 3 heardtones E, F, F#, G, G#, and A as

higher and the other tones aslower. So for this subject, F# and Gdefined the highest position alongthe pitch class circle. In contrast,the subject whose data are shownon the lower right heard tones A#,B, C, C#, D, and D# as higher andthe others as lower. So for this subject, C and C# defined the highestposition along the circle instead.Figure 4 shows the orientations ofthe pitch class circle for these 2 subjects (derived from the data shownin Fig. 3). I refer to tones that standat the top of a subject's pitch classcircle as his or her peak pitch classes.What can be the basis for thiscurious association between pitchclass and perceived height, and forthe individual differences in its

manifestation? In one experiment(Deutsch et al, 1987), the effect occurred in a substantial majority of a

group of undergraduates at theUniversity of California, San Diego. The subjects in this study allhad normal hearing and couldjudge correctly whether pairs ofpure tones that were separated byhalf-octaves formed ascending ordescending patterns. Within thispopulation, the way the tritone patterns were perceived did not correlate with whether or not the subjecthad musical training, so it appearsthat the tritone paradox is not musical in origin. Results from otherstudies argued against explanations in terms of low-level characteristics of the hearing mechanism.For example, the profiles relatingpitch class to perceived height inthe study of the tritone paradoxdid not correspond to patterns ofrelative loudness for the components of the same tones when thesewere compared with each other individually.

EVIDENCE FOR ANINFLUENCEOFSPEECH PATTERNSA number of informal observa

tions led me to conjecture that thetritone paradox might be related tothe processing of speech sounds.Specifically, I hypothesized thatthe listener develops a long-termrepresentation of the pitch range ofhis or her speaking voice, and thatthis representation includes a definition of the octave band inwhichthe largest proportion of pitch values occurs. I further hypothesizedthat the pitch classes delimitingthis octave band for speech aretaken by the listener as defining thehighest position along the pitchclass circle, thereby determiningthe way he or she orients the pitchclass circle with respect to height.

Together with Tom North andLee Ray, Iundertook a study to examine this hypothesis. First, wehad subjects participate in a full ex

o2O2LUO(/>LUOODC


6/8

178 VOLUME 6, NUMBER 6,DECEMBER 1997

! I_ __^

O^/c^n^L^(bj ^\v)& ?? Ce)

Fig. 4. Orientations of the pitch class circle with respect to height, for 2 differentsubjects. The circle on the left is derived from the data on the upper right of Figure3,with peak pitch classes of F# and G. The circle on the right is derived from the datashown on the lower right of Figure 3, with peak pitch classes of C and C#.

p?riment on the tritone paradox,and we selected a group whosejudgments showed clear and consistent relationships between pitchclass and perceived height. Thenwe had each subject speak freelyinto a microphone for roughly 15

min, and we took pitch estimates ofthe subject's speech samples at 4ms intervals. We then determinedthe octave in which the largestnumber of pitch estimates occurredand identified the pitch classes thatdefined the boundaries of this octave band. We then compared, foreach subject, the pitch classes delimiting the octave band for speechand those defining the highest position along the pitch class circle, asdetermined by that subject's judgments of the tritone paradox. For 8of the 9 subjects, the two sets of

pitch classes corresponded closely(Deutsch et al, 1990).The findings from this study arein accordance with the conjecturethat perception of the tritone para

dox is based on a representation ofthe pitch class circle by the listener,whose orientation is related to thepitch range of his or her speakingvoice. Two versions of this conjecture can then be considered. Thefirst, and more restricted, versiondoes not assume that the listener'spitch range for speech is itself determined by a learned template,but rather assumes it is determined

by some other factor. The second,and broader, version assumes thatsuch a template is acquiredthrough exposure to speech produced by other people, and that thetemplate is used both to evaluate

perceived speech and to constrainthe listener's own speech output.The characteristics of such a template would then be expected tovary among people who speak different languages or dialects, muchas other speech characteristics,such as vowel quality, vary acrosslanguages and dialects.A further study supported thesecond conjecture. Ihad noticed informally that Californians andpeople from the south of Englandtended to hear the tritone paradoxin different ways. In a formal experiment (Deutsch, 1991, 1994), Ifound that there was indeed a significant difference in the distribution of peak pitch classes between agroup of subjects who had grownup in California and a group whohad grown up in the south of England. As shown in Figure 5, theCalifornians tended to have peakpitch classes in the range B, C, C#,D, and D#, whereas the Englishgroup showed a different pattern,with the most frequent peak pitchclasses being F#, G, and G#.This experiment supports theview that through a learning process, an individual acquires a rep

J English30

20

10??03OCl(3o 0O A#B CC#DD# E FF#GG#ACO^-*COoo 30Q. Califomian

ii mu

2010H

XI AlA#B CC#DD# E FF#GG#A

pitch classFig. 5. Distributions of peak pitchclasses in two groups of listeners, onefrom the south of England and theother from California. The percentageof listeners forwhom each pitch class isa peak pitch class is graphed. FromDeutsch (1991).

resentation of the pitch class circlethat has a particular orientationwith respect to height. This orientation is derived from the speech towhich the person is exposed, andvaries from one language or dialectto another. Further, we can assumethat this template is involved bothin the person's own speech outputand in his or her evaluation ofspeech produced by others.Dolson (1994) has reviewed anumber of findings from thespeech literature that support thepresent conjecture. First, it appearsthat most people confine the pitchof their speech to a range ofroughly an octave. Second, withina given linguistic community, the



7/8

CURRENT DIRECTIONS INPSYCHOLOGICAL SCIENCE 179

speech of females is close to an octave above that of males. Third, thepitch ranges of speech differ re

markably little within a given linguistic community (except, ofcourse, for the gender difference).In contrast, there are considerabledifferences in the pitch ranges ofspeech across different linguisticcommunities. Finally, there is asurprising lack of correlation between the pitch range of an individual's speech and physiologicalparameters such as the person'sheight, weight, chest size, andlength of vocal tract.What could be the evolutionaryvalue of such a template? The pitchof a speaker's voice varies depending on his or her emotional state,and so serves to convey such statesto the listener (Fernald, 1992;Scherer, 1985). A template such asthis could provide a framework,common to a linguistic community,allowing the pitch of a speaker'svoice to be evaluated so that it cansignal his or her emotional state.The template could also be used inconveying syntactic aspects ofspeech (Cutler, Dahan, & Donselaar, 1997).

A number of other laboratorieshave recently produced additionalevidence supporting the idea thatthe orientation of the pitch classcircle varies statistically dependingon geographic community. Giangrande (1998), in a study of students at Florida Atlantic Universityin Boca Raton, Florida, obtained ahistogram of peak pitch classes thatwas quite similar to the one I obtained with Californians (Deutsch,1991). R. Treptoe (in unpublished

work) also produced a similar histogram from subjects at the University ofWisconsin, Steven's Point. Incontrast, Dawe, Platt, and Welsh(1998), in a study of students atMcMaster University in Hamilton,

Ontario, obtained a histogram thatwas very similar to the one I obtained from subjects from the southof England (Deutsch, 1991).

Further work from my laboratory has indicated that perceptionof the tritone paradox could wellbe influenced by a template that isformed

earlyin life.

Ragozzineand

I found statistical differences inperception of the tritone paradoxbetween two groups of subjectswho had all grown up in the areaof Youngstown, Ohio (Ragozzine& Deutsch, 1994). We designatedthose subjects whose parents hadalso grown up in Youngstown as"locals," and those whose parentshad grown up elsewhere as"aliens." The alien group produced a histogram of peak pitchclasses that was very similar to theone produced by my Californiansubjects (Deutsch, 1991), but the local group produced a very different histogram. Because parentshave a particularly strong influence on speech development, thisstudy indicates that an individual'spitch class template might beformed in childhood.

Most recently (Deutsch, 1996), Iobtained more direct evidence forsuch a developmentally acquiredtemplate. I studied the perceptionsof 15 subjects, together with thoseof their mothers. Ten of the subjectswere children, and 5 were adults.

The subjects were all Californian,but their mothers had grown up inmany different geographic regions,

including England, the Europeancontinent, and various parts of theUnited States. As expected, themothers perceived the tritone paradox in strikingly different ways.And remarkably, although all thesubjects were Californian, theirperceptions corresponded closelyto those of their mothers, and soalso differed considerably fromeach other. However, because 10 ofthe subjects were children, it is possible that the correlation betweenthe perceptions of children andtheir mothers reflects a particularlystrong maternal influence in thecase of young listeners?an issue

that is the subject of ongoing research.

CONCLUSION

Philosophers have argued forcenturies that strong linkages mustexist between music and speech.This view has been shared bymany composers who, in theirsearch for optimal expressivity,have incorporated into their musicfeatures that are characteristic ofspoken language. From a differentperspective, considerable advanceshave been made in documentingthe role of features based on pitchin the comprehension of spokenlanguage (Cutler et al., 1997) and inthe recognition of emotional state(Fernald, 1992; Scherer, 1985). The

present findings on the tritoneparadox indicate that, in addition,experience with speech can influence how music is perceived, andopen the door to uncovering othersuch influences.

Notes1. Address correspondence to Diana Deutsch, Department of Psychology, University of California, San Diego, La Jolla, CA 92093; e-mail:

[email protected]. A sine wave, or pure tone, is

composed of a single frequency. Thefrequency of a tone is the number oftimes in a second that the waveformrepeats itself. This is specified inHertz,where 1 Hertz (Hz) = 1 cycle per second. The amplitude of a tone is definedas the maximum amount of pressurevariation about the mean value. Forconvenience, amplitudes are convertedinto decibels (dB). The spectral envelope determines the amplitude of eachfrequency component.3. Signals comprising a full experiment on the tritone paradox, togetherwith a description of the proceduresfor testing subjects and analyzing thedata, are available in the followingcompact disk and accompanying booklet:Deutsch, D. (1995).Musical illusionsand paradoxes (Available from PhilomelRecords, P.O. Box 12189, La Jolla, CA92039-2189).

Copyright ? 1998 American Psychological Society


8/8

180 VOLUME 6, NUMBER 6,DECEMBER 1997

ReferencesBregman, A.S., & Campbell, J. (1971). Primary au

ditory stream segregation and perception oforder in rapid sequences of tones. Journal ofExperimental Psychology, 89, 244-249.

Cutler, A., Dahan, D., & Donselaar, W., van.(1997). Prosody in the comprehension of spoken language: A literature review. Languageand Speech, 40, 141-210.Dawe, L.A., Platt, J.R.,&Welsh, E. (1998). Spectralmotion after-effects and the tritone paradoxamong Canadian subjects. Perception & Psychophysics, 60, 209-220.Deutsch, D. (1975). Two-channel listening tomusical scales. Journal of the Acoustical Society ofAmerica, 57, 1156-1160.

Deutsch, D. (1986). A musical paradox. Music Perception, 3, 275-280.Deutsch, D. (1987). The tritone paradox: Effects ofspectral variables. Perception & Psychophysics,41, 563-575.

Deutsch, D. (1991). The tritone paradox: An influence of language on music perception. MusicPerception, 8, 335-347.Deutsch, D. (1994). The tritone paradox: Some further geographical correlates. Music Perception,12, 125-136.

Deutsch, D. (1996). Mothers and their childrenhear a musical illusion in strikingly similarways. Journal of the Acoustical Society ofAmerica, 99, 2482.

Deutsch, D., Kuyper, W.L., & Fisher, Y. (1987). Thetritone paradox: Its presence and form of distribution in a general population. Music Perception, 5, 79-92.

Deutsch, D., North, T., & Ray, L. (1990). The tritone paradox: Correlate with the listener's vocal range for speech. Music Perception, 7, 371384.Dolson, M. (1994). The pitch of speech as a function of linguistic community. Music Perception,11, 321-331.Fernald, A. (1992). Human maternal vocalizationsto infants as biologically relevant signals: An

evolutionary perspective. In J.H. Barkow, L.Cosmides, & J.Tooby (Eds.), The adapted mind:Evolutionary psychology and thegeneration of culture (pp. 391-427). New York: Oxford University Press.

Giangrande, J. (1998). The tritone paradox: Effectsof pitch class and position of the spectral envelope. Music Perception, 15, 253-264.

Hunt, F.V. (1978). Origins in acoustics. New Haven,CT: Yale University Press.

Penrose, L.S., & Penrose, R. (1958). Impossible ob

jects: A special type of illusion. British Journalof Psychology, 49, 31-33.

Ragozzine, F., & Deutsch, D. (1994). A regionaldifference in perception of the tritone paradoxwithin the United States. Music Perception, 12,213-225.Risset, J.C (1971). Paradoxes de hauteur. Proceed

ings of the 7th International Congree ofAcoustics,3, 613-616.Rock, I. (1986). The description and analysis of ob

ject and event perception. In K.R. Boff, L.Kaufman, & J.P. Thomas (Eds.), Handbook ofperception and human performance (chap. 33).New York: Wiley.Scherer, K.R. (1985). Vocal affect signaling: A comparative approach. Advances in the Study of Behavior, 15, 189-244.

Schouten, J.F. (1940). The perception of pitch. Philips Technical Review, 5, 286-294.

Seebeck, A. (1843). Ueber die Sirene. Annalen f?rPhysik und Chemie, 60, 449-481.

Shepard, R.N. (1964). Circularity in judgments ofrelative pitch. Journal of theAcoustical Society ofAmerica, 36, 2346-2353.Van Noorden, L.P.A.S. (1975). Temporal coherencein the perception of tone sequences. Unpublished

doctoral dissertation, Techniche Hogeschoel,Eindhoven, The Netherlands.

Animal Models Reveal the "Psych" inthe Psychosomatics of Peptic UlcersJ.Bruce Overmier and Robert Murison1Psychology Department, University of Minnesota, Minneapolis, Minnesota(J.B.O.), and Department of Biological & Medical Psychology, University ofBergen, Bergen, Norway (R.M.)

Ulcers have been regarded formany decades as the prototypicalpsychosomatic disease in which

psychological strain ("stress")leads to serious erosion and perforation of the stomach wall ("ulcers' ') Yet limited empirical evidence for stress as a causal factor inulcers (until recently) and recentclaims that they may be caused bybacterial infection have cast doubtson the psychosomatic view of ulcers. In this article, we examine evidence for the continued validity ofthe psychosomatic view of ulcers.

HISTORICAL CONTEXT

Selye (1936) introduced the concept of "stress," including amongits sources both biological and psy

chological factors and among itsconsequences peptic ulcers in boththe stomach (gastrum) and the entrance to the small intestine (duodenum). Selye believed ulcers arecaused by many factors and characterized them as "pluricausal."Both the psychological and the biological causes have remained a

mystery to this day, despite continued study by physicians, physiologists, and psychologists.The idea that psychological factors might cause physiological disease was not new with Selye. Indeed, the term "psychosomatic,"always much criticized, can betraced to the 19th century. Mesmer,Charcot, Janet, and Freud exploited"psychological

treatments" forphysical symptoms. In U.S. popular culture, peptic ulcer became aprototype for psychosomatic disease. Wolf and Wolff (1947) directly observed the functioning ofthe stomach in a patient calledTom; they observed that the stomach responds to psychological challenges. These classic observationsconfirmed that psychological factors can influence the gastrointesti

Recommended ReadingGlavin, G.B., Murison, R., Overmier,

J.B., Par?, W.P., Bakke, H.K.,Henke, P.G., & Hernandez, D.E.

(1991). The neurobiology of stressulcers. Brain Research Reviews, 16,301-343.

Sapolsky, RM. (1994). Why zebrasdon't get ulcers. San Francisco:W.H. Freeman.

Weiner, H. (1992). Perturbing the organism: The biology of stressful experience. Chicago: University of Chicago Press.Wolf, S. (1981). The psyche and thestomach: A historical vignette.Gastroenterology, 80, 605-614.


the tritone paradox: a link between music and speech

Documents