This article was downloaded by: [University of Texas Libraries] On: 04 December 2014, At: 16:53 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Quarterly Journal of Experimental Psychology Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/pqje19
Visual perception of musical notation: Registering pitch symbols in memory John A. Sloboda
a
a
Department of Psychology , University of Keele , Staffordshire, ST5 5BG, U.K. Published online: 29 May 2007.
To cite this article: John A. Sloboda (1976) Visual perception of musical notation: Registering pitch symbols in memory, Quarterly Journal of Experimental Psychology, 28:1, 1-16, DOI: 10.1080/14640747608400532 To link to this article: http://dx.doi.org/10.1080/14640747608400532
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden.
Downloaded by [University of Texas Libraries] at 16:53 04 December 2014
Terms & Conditions of access and use can be found at http://www.tandfonline.com/ page/terms-and-conditions
Quarterly Journal of Experimental Psychology
( I 976)
28, 1-16
VISUAL PERCEPTION OF MUSICAL NOTATION : REGISTERING PITCH SYMBOLS IN MEMORY Downloaded by [University of Texas Libraries] at 16:53 04 December 2014
JOHN A. SLOBODA Department of Psychology, University of Keele, Stafordshire ST5 5BG, U.K . Music bears formal relations to language which suggest that perceptual processes in the two modes may also be similar. These experiments examined the way in which experienced musicians differed from non-musicians in their recognition of briefly exposed pitch notation. Experiments I and I1 together demonstrated that musicians are superior to non-musicians in their immediate written recall of stimuli containing more than three notes, but only when the stimulus is available to them for 150 ms or more. These results are accounted for well by a model proposed by Coltheart (1972) for letter perception under conditions of brief exposure. In this model, two coding processes act simultaneously on the stimulus, one a fast visual coding, and the other a slower, but more permanent abstract (or name) coding. In this case non-musicians appear to be lacking a second, abstract, coding which musicians possess. Experiments I11 and IV attempted to investigate the nature of the abstract code for musicians by presenting various types of interference in the linguistic or musical mode. Neither concurrent letter naming nor concurrent memorization of pitches appeared to cause a decrement in the original visual task, suggesting that musicians may not have been using simple naming or pitching transformations in coding the visual input.
Introduction T h e way in which visual symbols are recognized depends not only upon their visual form but also upon their function within the system that they symbolize. Research into the perception of letters, the visual symbols of language, has suggested that the reader learns and makes automatic a hierarchy of coding processes which transform input into progressively more abstract form (Gough, 1972; LaBerge and Samuels, 1974). Thus, at the simplest level each letter may be named (Sperling, 1967; Coltheart, 1972), but recognition can also be assisted if letters are contained in pronounceable groups (Gibson, Pick, Osser and Hammond, 1962; Spoehr and Smith, 1973) or in words (Reicher, 1969; Wheeler, 1970). There is also evidence that the perceptual processes in reading are sensitive to phrase structure (Levin and Kaplan, 1970). Studies of non-linguistic symbol systems have been less systematic, although some recent experiments have suggested that knowledge of the abstract structure {behindthe symbols can affect their perception in a manner analogous to language. For instance, Baron and Thurston (1973) have demonstrated a kind of Word Superiority Effect (Reicher, 1969) for chemical symbols. They showed that trained chemists are better at identifying conventionally notated compounds than compounds in which the order of anion and cation have been reversed. Studies of I
Downloaded by [University of Texas Libraries] at 16:53 04 December 2014
2
J. A. SLOBODA
visual memory in chess (Chase and Simon, 1973) have shown large differences in the ability of chess players of different standards to reproduce a briefly exposed position from a game. These differences disappear when the stimulus is a random array of pieces on a board. Simon and Gilmartin (1973) have used the EPAM model of verbal information processing to accurately simulate the performance of chess players in this situation. Reicher (personal communication) has studied visual memory for musical notation in musicians of various standards, and suggests mechanisms of music perception analogous to those operating in chess perception. The similarity of music to speech has been noted in a number of contexts. Winograd (1968) has demonstrated that one may approach the analysis of tonal harmony in much the same way as the analysis of grammatical structure in language. In addition there are clear behavioural analogies between language and music (Gates, Bradshaw, and Nettleton, 1974; Sloboda, 1974~2):both are composed of sequences of discrete sounds; vocal production and auditory reception are primary in both; and both use arbitrary visual symbols to notate salient aspects of the sound pattern. Also, reading of text at speed requires many years to develop in both skills, and it is reasonable to suppose that the music reader, like the language reader, increases his coding efficiency through attention to structures in the text (Sloboda, I 9-74) The simplest kind of coding operation is the categorization of an individual symbol by, for instance, naming it; and the experiments to follow were designed to investigate such an operation in the perceptual performance of musicians. There are many studies which show that naming takes place automatically when experienced readers perform perceptual tasks involving letters (Posner, Lewis and Conrad, 1972; Keele, 1972; Shiffrin and Gardner, 1972). Equally, visual symbols with no names are often poorly coded by subjects (Grindley and Townsend, 19-73). Coltheart (1972) has suggested that experienced readers initiate two coding processes when presented with an array of letters. One is a visual coding which takes about 20 ms per letter and which can store up to four items. The other is a naming code which takes about IOO ms per letter and which can store up to seven or eight items. In normal circumstances information from the visual code decays rapidly whilst the name code is retained. Stimulus exposures of less than IOO ms, however, seem to preclude the operation of the naming code. A musical symbol offers similar opportunity for dual coding. Not only is it visually defined but it represents a note with an alphabetical name and a certain musical pitch. If we assume that non-visual coding takes longer than visual coding then we may examine the existence of dual coding in musicians by observing perceptual performance at various exposure durations of a musical stimulus. Do musicians differ from nonmusicians through a better-developed visual coding mechanism, or is their perceptual performance governed by their ability to translate visual information into non-visual name-like codes ? If musicians have better visual codes then their ability to report individual notes presented very briefly should be superior to nonmusicians. If, however, they differ from non-musicians in possessing welldeveloped nonvisual codes then their superior ability should only be present for stimuli which are presented at durations long enough to allow these codes to come into operation.
Downloaded by [University of Texas Libraries] at 16:53 04 December 2014
MEMORY FOR PITCH SYMBOLS
3
Experiment I Bean (1938) first investigated performance of musicians under tachistoscopic conditions of exposure, and found that good readers could report more notes from a display than could poor readers. His measure of performance, however, was to have subjects play what they saw on a piano. It is not, therefore, absolutely clear how much of the superiority was due to perceptual factors and how much due to factors of manual response. I n the present experiment subjects were required to write down what they could see of a simple musical stimulus in which a sequence of up to six pitch symbols (without accidentals, rhythmic "tails", or bar lines) was written on a single stave. This eliminated the possibility that instrumental competence would give some subjects a response advantage, and the artificially simple stimuli were so unlike normal musical passages that experienced musicians would not be likely to recognize familiar themes or patterns in what they saw. T h e experiment examined performance of musicians and non-musicians under two conditions of exposure, a 20-ms stimulus duration, and a 2-s stimulus duration,
Method Materials Six by 4 " (152 x 102 mm) cards were prepared on each of which was printed a musical stave of 5 stave lines extending across the whole card. The vertical distance between stave lines was 3/16" (4.8 mm), and there was a single vertical fixation line 14" (43 m) to the left of centre. Note arrays of varying lengths were written onto these cards, the leftmost note always being centred 3" (9.5 mm) to the right of the fixation line, and each subsequent note centred another t m (9.5 mm) to the right. The notes themselves were filled black circles of diameter &" (3.2 mm) without tails. A note array therefore had no rhythmic implication, being in the form commonly used to indicate a freely flowing melodic chant. 96 cards were prepared altogether, 16 for each array length from I to 6. These were constructed from random number tables with the following constraints: no array should contain the same note twice; and each possible leftmost note should be used approximately equal numbers of times. The experimental subset of possible notes was in fact 10, the uppermost note above the stave system never being employed. The subject was provided with a response sheet containing staves of the same size as the stimulus staves. The stimuli were presented in a standard Cambridge two-field tachistoscope. The adaptation stimulus was a stimulus card containing a stave and a fixation line, so positioned that it superimposed the stimulus stave. The subjective impression of a stimulus exposure was thus of notes appearing on a permanently present stave. Figure I shows an example of a stimulus card.
I FIGURE I . Example of stimulus used in Experiment I.
4
J. A. SLOBODA
Downloaded by [University of Texas Libraries] at 16:53 04 December 2014
Subjects Eight subjects took part in the experiment. Four of these were undergraduates of London University who had received little or no musical training and were naive with respect to musical notation. The other four subjects were music students who were highly experienced in dealing with musical notation. Subjects had normal or corrected vision and the musicians were all keyboard instrumentalists with at least 10years regular experience of music reading.
Procedure The subject was instructed that the experiment was concerned with accuracy of visual perception for music-like material. He was therefore to observe each stimulus carefully and immediately following offset to attempt to record the exact sequence of notes (i.e. in the correct left-right order) on the response sheet. Each subject received 10 practice trials during which it was determined that the stimuli were clearly visible, and that the subject knew what the response alternatives were, and how to indicate his choice rapidly and efficiently (i.e. this involved the knowledge, for non-musicians, that the note either covered a stave line or fell in between two lines. Intermediate positions were never used and had no musical significance). The 96 stimulus cards were presented in random order and split between two experimental conditions, 20-ms exposure duration and 2-s exposure duration, which were distributed throughout the experimental sequence. On each exposure the subject was informed of the number of notes in each stimulus in advance, and was required to produce the same number of notes in his response. The subject initiated his own exposures by means of a thumbswitch held in the non-writing hand. Trials were administered in blocks of 24 with rest pauses of about I min between blocks.
Results Responses were initially scored according to whether a note in a particular position was correct or not, and the total score for each array was computed. This score was subject to an analysis of variance taking into account subject group, subjects within groups, exposure duration, and array length. T h e mean scores for each group are shown in Table I. Musicians were significantly better than non-musicians at the task ( F = 12.61,df = I , I , P
Quarterly Journal of Experimental Psychology Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/pqje19
Visual perception of musical notation: Registering pitch symbols in memory John A. Sloboda
a
a
Department of Psychology , University of Keele , Staffordshire, ST5 5BG, U.K. Published online: 29 May 2007.
To cite this article: John A. Sloboda (1976) Visual perception of musical notation: Registering pitch symbols in memory, Quarterly Journal of Experimental Psychology, 28:1, 1-16, DOI: 10.1080/14640747608400532 To link to this article: http://dx.doi.org/10.1080/14640747608400532
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden.
Downloaded by [University of Texas Libraries] at 16:53 04 December 2014
Terms & Conditions of access and use can be found at http://www.tandfonline.com/ page/terms-and-conditions
Quarterly Journal of Experimental Psychology
( I 976)
28, 1-16
VISUAL PERCEPTION OF MUSICAL NOTATION : REGISTERING PITCH SYMBOLS IN MEMORY Downloaded by [University of Texas Libraries] at 16:53 04 December 2014
JOHN A. SLOBODA Department of Psychology, University of Keele, Stafordshire ST5 5BG, U.K . Music bears formal relations to language which suggest that perceptual processes in the two modes may also be similar. These experiments examined the way in which experienced musicians differed from non-musicians in their recognition of briefly exposed pitch notation. Experiments I and I1 together demonstrated that musicians are superior to non-musicians in their immediate written recall of stimuli containing more than three notes, but only when the stimulus is available to them for 150 ms or more. These results are accounted for well by a model proposed by Coltheart (1972) for letter perception under conditions of brief exposure. In this model, two coding processes act simultaneously on the stimulus, one a fast visual coding, and the other a slower, but more permanent abstract (or name) coding. In this case non-musicians appear to be lacking a second, abstract, coding which musicians possess. Experiments I11 and IV attempted to investigate the nature of the abstract code for musicians by presenting various types of interference in the linguistic or musical mode. Neither concurrent letter naming nor concurrent memorization of pitches appeared to cause a decrement in the original visual task, suggesting that musicians may not have been using simple naming or pitching transformations in coding the visual input.
Introduction T h e way in which visual symbols are recognized depends not only upon their visual form but also upon their function within the system that they symbolize. Research into the perception of letters, the visual symbols of language, has suggested that the reader learns and makes automatic a hierarchy of coding processes which transform input into progressively more abstract form (Gough, 1972; LaBerge and Samuels, 1974). Thus, at the simplest level each letter may be named (Sperling, 1967; Coltheart, 1972), but recognition can also be assisted if letters are contained in pronounceable groups (Gibson, Pick, Osser and Hammond, 1962; Spoehr and Smith, 1973) or in words (Reicher, 1969; Wheeler, 1970). There is also evidence that the perceptual processes in reading are sensitive to phrase structure (Levin and Kaplan, 1970). Studies of non-linguistic symbol systems have been less systematic, although some recent experiments have suggested that knowledge of the abstract structure {behindthe symbols can affect their perception in a manner analogous to language. For instance, Baron and Thurston (1973) have demonstrated a kind of Word Superiority Effect (Reicher, 1969) for chemical symbols. They showed that trained chemists are better at identifying conventionally notated compounds than compounds in which the order of anion and cation have been reversed. Studies of I
Downloaded by [University of Texas Libraries] at 16:53 04 December 2014
2
J. A. SLOBODA
visual memory in chess (Chase and Simon, 1973) have shown large differences in the ability of chess players of different standards to reproduce a briefly exposed position from a game. These differences disappear when the stimulus is a random array of pieces on a board. Simon and Gilmartin (1973) have used the EPAM model of verbal information processing to accurately simulate the performance of chess players in this situation. Reicher (personal communication) has studied visual memory for musical notation in musicians of various standards, and suggests mechanisms of music perception analogous to those operating in chess perception. The similarity of music to speech has been noted in a number of contexts. Winograd (1968) has demonstrated that one may approach the analysis of tonal harmony in much the same way as the analysis of grammatical structure in language. In addition there are clear behavioural analogies between language and music (Gates, Bradshaw, and Nettleton, 1974; Sloboda, 1974~2):both are composed of sequences of discrete sounds; vocal production and auditory reception are primary in both; and both use arbitrary visual symbols to notate salient aspects of the sound pattern. Also, reading of text at speed requires many years to develop in both skills, and it is reasonable to suppose that the music reader, like the language reader, increases his coding efficiency through attention to structures in the text (Sloboda, I 9-74) The simplest kind of coding operation is the categorization of an individual symbol by, for instance, naming it; and the experiments to follow were designed to investigate such an operation in the perceptual performance of musicians. There are many studies which show that naming takes place automatically when experienced readers perform perceptual tasks involving letters (Posner, Lewis and Conrad, 1972; Keele, 1972; Shiffrin and Gardner, 1972). Equally, visual symbols with no names are often poorly coded by subjects (Grindley and Townsend, 19-73). Coltheart (1972) has suggested that experienced readers initiate two coding processes when presented with an array of letters. One is a visual coding which takes about 20 ms per letter and which can store up to four items. The other is a naming code which takes about IOO ms per letter and which can store up to seven or eight items. In normal circumstances information from the visual code decays rapidly whilst the name code is retained. Stimulus exposures of less than IOO ms, however, seem to preclude the operation of the naming code. A musical symbol offers similar opportunity for dual coding. Not only is it visually defined but it represents a note with an alphabetical name and a certain musical pitch. If we assume that non-visual coding takes longer than visual coding then we may examine the existence of dual coding in musicians by observing perceptual performance at various exposure durations of a musical stimulus. Do musicians differ from nonmusicians through a better-developed visual coding mechanism, or is their perceptual performance governed by their ability to translate visual information into non-visual name-like codes ? If musicians have better visual codes then their ability to report individual notes presented very briefly should be superior to nonmusicians. If, however, they differ from non-musicians in possessing welldeveloped nonvisual codes then their superior ability should only be present for stimuli which are presented at durations long enough to allow these codes to come into operation.
Downloaded by [University of Texas Libraries] at 16:53 04 December 2014
MEMORY FOR PITCH SYMBOLS
3
Experiment I Bean (1938) first investigated performance of musicians under tachistoscopic conditions of exposure, and found that good readers could report more notes from a display than could poor readers. His measure of performance, however, was to have subjects play what they saw on a piano. It is not, therefore, absolutely clear how much of the superiority was due to perceptual factors and how much due to factors of manual response. I n the present experiment subjects were required to write down what they could see of a simple musical stimulus in which a sequence of up to six pitch symbols (without accidentals, rhythmic "tails", or bar lines) was written on a single stave. This eliminated the possibility that instrumental competence would give some subjects a response advantage, and the artificially simple stimuli were so unlike normal musical passages that experienced musicians would not be likely to recognize familiar themes or patterns in what they saw. T h e experiment examined performance of musicians and non-musicians under two conditions of exposure, a 20-ms stimulus duration, and a 2-s stimulus duration,
Method Materials Six by 4 " (152 x 102 mm) cards were prepared on each of which was printed a musical stave of 5 stave lines extending across the whole card. The vertical distance between stave lines was 3/16" (4.8 mm), and there was a single vertical fixation line 14" (43 m) to the left of centre. Note arrays of varying lengths were written onto these cards, the leftmost note always being centred 3" (9.5 mm) to the right of the fixation line, and each subsequent note centred another t m (9.5 mm) to the right. The notes themselves were filled black circles of diameter &" (3.2 mm) without tails. A note array therefore had no rhythmic implication, being in the form commonly used to indicate a freely flowing melodic chant. 96 cards were prepared altogether, 16 for each array length from I to 6. These were constructed from random number tables with the following constraints: no array should contain the same note twice; and each possible leftmost note should be used approximately equal numbers of times. The experimental subset of possible notes was in fact 10, the uppermost note above the stave system never being employed. The subject was provided with a response sheet containing staves of the same size as the stimulus staves. The stimuli were presented in a standard Cambridge two-field tachistoscope. The adaptation stimulus was a stimulus card containing a stave and a fixation line, so positioned that it superimposed the stimulus stave. The subjective impression of a stimulus exposure was thus of notes appearing on a permanently present stave. Figure I shows an example of a stimulus card.
I FIGURE I . Example of stimulus used in Experiment I.
4
J. A. SLOBODA
Downloaded by [University of Texas Libraries] at 16:53 04 December 2014
Subjects Eight subjects took part in the experiment. Four of these were undergraduates of London University who had received little or no musical training and were naive with respect to musical notation. The other four subjects were music students who were highly experienced in dealing with musical notation. Subjects had normal or corrected vision and the musicians were all keyboard instrumentalists with at least 10years regular experience of music reading.
Procedure The subject was instructed that the experiment was concerned with accuracy of visual perception for music-like material. He was therefore to observe each stimulus carefully and immediately following offset to attempt to record the exact sequence of notes (i.e. in the correct left-right order) on the response sheet. Each subject received 10 practice trials during which it was determined that the stimuli were clearly visible, and that the subject knew what the response alternatives were, and how to indicate his choice rapidly and efficiently (i.e. this involved the knowledge, for non-musicians, that the note either covered a stave line or fell in between two lines. Intermediate positions were never used and had no musical significance). The 96 stimulus cards were presented in random order and split between two experimental conditions, 20-ms exposure duration and 2-s exposure duration, which were distributed throughout the experimental sequence. On each exposure the subject was informed of the number of notes in each stimulus in advance, and was required to produce the same number of notes in his response. The subject initiated his own exposures by means of a thumbswitch held in the non-writing hand. Trials were administered in blocks of 24 with rest pauses of about I min between blocks.
Results Responses were initially scored according to whether a note in a particular position was correct or not, and the total score for each array was computed. This score was subject to an analysis of variance taking into account subject group, subjects within groups, exposure duration, and array length. T h e mean scores for each group are shown in Table I. Musicians were significantly better than non-musicians at the task ( F = 12.61,df = I , I , P
