- Email: [email protected]
Contrasting conceptions of human infants
Research Update
326
TRENDS in Cognitive Sciences Vol.6 No.8 August 2002
Meeting Report
Contrasting conceptions of human infants Sandra E. Trehub The XIII Biennial International Conference on Infant Studies was held in Toronto, Canada on 18–21 April 2002.
The conference was wide-ranging, and included keynote addresses, debates, symposia, discussions and poster presentations on current issues in infancy research. Attachment and culture
In a debate on the universal relevance of attachment theory, Fred Rothbaum (Tufts University, Medford, MA, USA) argued that current views are ethnocentric and originate from Western values such as autonomy, self-expression and exploration. Thus, they are largely irrelevant to Eastern cultures (and to Western subcultures) that value interdependence and harmony. Rothbaum noted that identical parental behaviours can have positive connotations in one cultural milieu and negative connotations in another. For example, the extent of parental directiveness that is highly valued in Korea, China and elsewhere would be considered overly controlling in middle-class Western contexts. Responsiveness to infant signals, which is considered a marker of caregiver sensitivity, is also culturespecific. For example, Western mothers are more responsive to infants’ non-distress vocalizations (e.g. cooing and babbling) than are non-Western mothers, but are less responsive to infants’ distress vocalizations (e.g. crying and fussing). These and other crosscultural differences have led Rothbaum and his colleagues to question the generality of attachment theory, the processes presumed to underlie attachment behaviours, and the usual methods of study [1]. Everett Waters (State University of New York, Stony Brook, NY, USA) considered whether the cross-cultural differences undermined the basic tenets of classic attachment theory, as articulated by John Bowlby [2]. Bowlby conceptualized the mother–infant bond as a secure-base relationship. It follows, from his http://tics.trends.com
conceptualization, that infants with available, supportive caregivers would be more motivated to explore their environment than would infants with less available or less consistent caregivers. Bowlby believed that the caregiving environment, operating in conjunction with infants’ inherent learning biases, fostered the development of a system of attachment behaviours. Presumably, differences in caregiving experience would generate differences in the attachment control system, specifically in infants’ use of the relationship for achieving their own goals. Waters argued against tests of Bowlby’s theory that depend upon identical caregiving or infant behaviours across cultures. The crucial tests should focus on similar underlying principles – caregivers’ sensitivity to infant signals, their continuing supervision and availability, and infants’ use of their caregiver as a secure base [3]. He acknowledged that laboratory tests of infant attachment are meaningful only to the extent that they elicit infants’ characteristic use of their caregiver as a secure base. Waters also argued that differences in the proportion of ‘securely attached’ infants across cultures pose no threat to Bowlby’s theory, because differences in caregivers’ availability and responsiveness (i.e. infants’ prior experience) might account for such discrepancies. Instead of viewing the cross-cultural findings as a threat to classic attachment theory, Waters considered them an opportunity to preserve, extend or revise its key tenets, which emerged from Bowlby’s reconsideration of Freud’s ideas about the mother–infant bond. Music processing in infancy
Increasing consideration of the evolutionary origins of music [4] is generating interdisciplinary interest in naïve listeners’ perception of music. The presumption is that nature’s contribution should be most readily apparent in infancy, before culture has left its indelible mark. A number of papers at this meeting examined infants’ memory for musical
patterns and musical preferences. Beatriz Ilari (McGill University, Montreal, QC, Canada) provided 8-month-old infants with ten days of limited exposure to one of two musical selections (piano pieces by Maurice Ravel) that were considerably more complex than those used previously with such young listeners. Infants preferred the familiar to the unfamiliar piano piece when tested two weeks after the exposure period. Laurel Trainor (McMaster University, Hamilton, ON, Canada) and colleagues found that seven days of exposure to piano renditions of an unfamiliar folk tune (three minutes each day) enabled 6-month-old infants to distinguish that tune from another when tested one week after the exposure phase. When a harp played the same tunes, infants retained them for longer periods, which implies that a distinctive timbre supports a unique encoding and a more enduring memory trace. Alterations of the timbre (by changing the instrument) and tempo (speed) between exposure and test disrupted retention, even without a delay between familiarization and test. Thus, infants’ encoding of melodies seems to be feature-specific, at least for timbre and tempo. Infants recognized transpositions of the tunes to a new pitch range (i.e. novel pitches but identical pitch relations), which indicates that they encode melodies in terms of relative pitch. This finding contradicts previous claims of preferential processing for absolute over relative pitch in infancy [5]. Hiromi Nito (Yamaha Music Foundation, Tokyo, Japan) and associates examined age-related changes in infants’ processing of tunes that are commonly sung during Japanese childhood. When infants were given a choice of tunes in their original mode (major key) or in a novel mode (parallel minor key), 5- to 11-month-olds showed no preference, but 14-month-olds preferred the novel (minor) to the familiar (major) mode. It is unclear whether the younger infants recognized the familiar tunes and whether they could distinguish major from minor renditions.
1364-6613/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S1364-6613(02)01944-7
Research Update
If the encoding of tunes is timbre-specific, as Trainor and her associates suggest, then young infants could have failed to connect the woodwind renditions in Hito’s study with the vocal renditions of everyday life. Trehub and colleagues (University of Toronto at Mississauga, Mississauga, ON, Canada) presented research on the relation between musical preferences and non-musical behaviours. When 6- and 9-month-old infants were provided with a choice of soothing or playful performances of the same song by the same singer, their preferences were predictable from their
TRENDS in Cognitive Sciences Vol.6 No.8 August 2002
patterns of visual fixation. Specifically, infants with extended visual fixations (i.e. slow visual tempo) preferred the soothing performances (much slower tempo), and infants with consistently brief visual fixations preferred the playful performances (a faster tempo). These studies of music processing in infancy will contribute to the ongoing debate about the biological and cultural foundations of music. References 1 Rothbaum, F. et al. (2000) Attachment and culture: security in the United States and Japan. Am. Psychol. 55, 1093–1104
327
2 Bowlby, J. (1969) Attachment and Loss: Vol 1. Attachment, Basic Books 3 Waters, E. and Cummings, E.M. (2000) A secure base from which to explore close relationships. Child Devel. 71, 164–172 4 Wallin, N. et al. (2000) The Origins of Music, MIT Press 5 Saffran, J.R. and Griepentrog, G.J. (2001) Absolute pitch in infant auditory learning: evidence of developmental reorganization. Dev. Psychol. 37, 74–85
Sandra E. Trehub Dept of Psychology, University of Toronto at Mississauga, Mississauga, Canada L5L 1C6. e-mail: [email protected]
Making connections about brain connectivity James V. Stone and Rolf Kötter The Functional Brain Connectivity workshop was organized by Rolf Kötter and Karl Friston, and held in Düsseldorf, Germany, on 4–6 April, 2002.
This three-day workshop on brain connectivity comprised 15 presentations by speakers using a wide variety different research techniques, including functional magnetic resonance imaging (fMRI), diffusion tensor imaging (DTI), optical imaging (OI) and computational modelling. Each speaker used only two transparencies, resulting in a very interactive and informal workshop. As might be expected, the resultant animated discussions between researchers from different scientific disciplines generated much heat, but also a few beams of much needed light. Functional, effective and structural connectivity
The very first presentation by Wim Vanduffel (Leuven Medical School, Belgium) led to a defining theme of the workshop: the distinction between functional, effective and structural connectivity. Essentially, structural connectivity refers to the direct anatomical connections between brain regions, whereas functional connectivity refers to a statistical dependency between the activities of different regions without regard to any underlying anatomical connection. These terms were coined during the pioneering attempts by Gerstein (University of Pennsylvania, Philadelphia, http://tics.trends.com
PA, USA), Aertsen (University of Freiburg, Freiburg, Germany) and colleagues to infer the nature of neuronal interactions from multi-electrode recordings, as recalled by Günther Palm (University of Ulm, Germany). Building on these early discussions, the distinction between functional and effective connectivity was first formally identified by Karl Friston (Wellcome Department of Cognitive Neurology, London, UK) in 1994. With Friston on hand to resolve disagreements over these definitions, it emerged that his original motivation for making such a distinction was to force researchers to be explicit about the nature of putative relationships among brain regions. Specifically, Friston argued that establishing effective connectivity between brain regions demands a precise mathematical specification of the nature of the proposed functional relationship in a way that functional connectivity does not. In other words, effective connectivity implies a precise mathematical model, whereas functional connectivity does not. The ensuing discussion soon showed that this seemingly clear distinction leads on to the question of what characterizes a model. Analyzing connectivity
Presentations of methods for analysing connectivity between brain regions were divided into two groups: those investigating structural connectivity using in vivo DTI (Rainer Goebel, University of Maastricht, The Netherlands; Martin A. Koch, Max Planck Institute of
Cognitive Neuroscience, Leipzig, Germany) and those using post-mortem human anatomical techniques (Karl Zilles, Research Centre Jülich, Germany). Owing to its technical limitations, DTI is presently no rival to in vivo tracer or post-mortem myeloarchitectonic methods. Indeed, Zilles emphasized that anatomical techniques imply that the Talairach atlas, which is the de facto standard for imaging coordinates, is essentially inaccurate. Zilles cogently made this point by presenting actual brain slices (which neatly subverted the conference limit of two transparencies). Despite the limitations of DTI, Goebel demonstrated the potential for this method to identify the anatomical pathways associated with activations observed using fMRI in the same subject. Methods for inferring connectivity included independent component analysis (James Stone, Sheffield University, UK), dynamic causal modelling (Karl Friston, Wellcome Functional Imaging Laboratory, London, UK), and structural equation modelling (Anthony Randy McIntosh, University of Toronto, Canada). A freeranging and provocative discussion on the role of causality in connectivity was also led by Ed Bullmore (University of Cambridge, UK). Presentations of methods for inferring the local functional topography of single cortical regions included optical imaging (Peter Buzás, Ruhr-Universität Bochum, Germany; Mark Hübener, Max-PlanckInstitut fur Neurobiologie, Martinsried,
1364-6613/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S1364-6613(02)01943-5
326
TRENDS in Cognitive Sciences Vol.6 No.8 August 2002
Meeting Report
Contrasting conceptions of human infants Sandra E. Trehub The XIII Biennial International Conference on Infant Studies was held in Toronto, Canada on 18–21 April 2002.
The conference was wide-ranging, and included keynote addresses, debates, symposia, discussions and poster presentations on current issues in infancy research. Attachment and culture
In a debate on the universal relevance of attachment theory, Fred Rothbaum (Tufts University, Medford, MA, USA) argued that current views are ethnocentric and originate from Western values such as autonomy, self-expression and exploration. Thus, they are largely irrelevant to Eastern cultures (and to Western subcultures) that value interdependence and harmony. Rothbaum noted that identical parental behaviours can have positive connotations in one cultural milieu and negative connotations in another. For example, the extent of parental directiveness that is highly valued in Korea, China and elsewhere would be considered overly controlling in middle-class Western contexts. Responsiveness to infant signals, which is considered a marker of caregiver sensitivity, is also culturespecific. For example, Western mothers are more responsive to infants’ non-distress vocalizations (e.g. cooing and babbling) than are non-Western mothers, but are less responsive to infants’ distress vocalizations (e.g. crying and fussing). These and other crosscultural differences have led Rothbaum and his colleagues to question the generality of attachment theory, the processes presumed to underlie attachment behaviours, and the usual methods of study [1]. Everett Waters (State University of New York, Stony Brook, NY, USA) considered whether the cross-cultural differences undermined the basic tenets of classic attachment theory, as articulated by John Bowlby [2]. Bowlby conceptualized the mother–infant bond as a secure-base relationship. It follows, from his http://tics.trends.com
conceptualization, that infants with available, supportive caregivers would be more motivated to explore their environment than would infants with less available or less consistent caregivers. Bowlby believed that the caregiving environment, operating in conjunction with infants’ inherent learning biases, fostered the development of a system of attachment behaviours. Presumably, differences in caregiving experience would generate differences in the attachment control system, specifically in infants’ use of the relationship for achieving their own goals. Waters argued against tests of Bowlby’s theory that depend upon identical caregiving or infant behaviours across cultures. The crucial tests should focus on similar underlying principles – caregivers’ sensitivity to infant signals, their continuing supervision and availability, and infants’ use of their caregiver as a secure base [3]. He acknowledged that laboratory tests of infant attachment are meaningful only to the extent that they elicit infants’ characteristic use of their caregiver as a secure base. Waters also argued that differences in the proportion of ‘securely attached’ infants across cultures pose no threat to Bowlby’s theory, because differences in caregivers’ availability and responsiveness (i.e. infants’ prior experience) might account for such discrepancies. Instead of viewing the cross-cultural findings as a threat to classic attachment theory, Waters considered them an opportunity to preserve, extend or revise its key tenets, which emerged from Bowlby’s reconsideration of Freud’s ideas about the mother–infant bond. Music processing in infancy
Increasing consideration of the evolutionary origins of music [4] is generating interdisciplinary interest in naïve listeners’ perception of music. The presumption is that nature’s contribution should be most readily apparent in infancy, before culture has left its indelible mark. A number of papers at this meeting examined infants’ memory for musical
patterns and musical preferences. Beatriz Ilari (McGill University, Montreal, QC, Canada) provided 8-month-old infants with ten days of limited exposure to one of two musical selections (piano pieces by Maurice Ravel) that were considerably more complex than those used previously with such young listeners. Infants preferred the familiar to the unfamiliar piano piece when tested two weeks after the exposure period. Laurel Trainor (McMaster University, Hamilton, ON, Canada) and colleagues found that seven days of exposure to piano renditions of an unfamiliar folk tune (three minutes each day) enabled 6-month-old infants to distinguish that tune from another when tested one week after the exposure phase. When a harp played the same tunes, infants retained them for longer periods, which implies that a distinctive timbre supports a unique encoding and a more enduring memory trace. Alterations of the timbre (by changing the instrument) and tempo (speed) between exposure and test disrupted retention, even without a delay between familiarization and test. Thus, infants’ encoding of melodies seems to be feature-specific, at least for timbre and tempo. Infants recognized transpositions of the tunes to a new pitch range (i.e. novel pitches but identical pitch relations), which indicates that they encode melodies in terms of relative pitch. This finding contradicts previous claims of preferential processing for absolute over relative pitch in infancy [5]. Hiromi Nito (Yamaha Music Foundation, Tokyo, Japan) and associates examined age-related changes in infants’ processing of tunes that are commonly sung during Japanese childhood. When infants were given a choice of tunes in their original mode (major key) or in a novel mode (parallel minor key), 5- to 11-month-olds showed no preference, but 14-month-olds preferred the novel (minor) to the familiar (major) mode. It is unclear whether the younger infants recognized the familiar tunes and whether they could distinguish major from minor renditions.
1364-6613/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S1364-6613(02)01944-7
Research Update
If the encoding of tunes is timbre-specific, as Trainor and her associates suggest, then young infants could have failed to connect the woodwind renditions in Hito’s study with the vocal renditions of everyday life. Trehub and colleagues (University of Toronto at Mississauga, Mississauga, ON, Canada) presented research on the relation between musical preferences and non-musical behaviours. When 6- and 9-month-old infants were provided with a choice of soothing or playful performances of the same song by the same singer, their preferences were predictable from their
TRENDS in Cognitive Sciences Vol.6 No.8 August 2002
patterns of visual fixation. Specifically, infants with extended visual fixations (i.e. slow visual tempo) preferred the soothing performances (much slower tempo), and infants with consistently brief visual fixations preferred the playful performances (a faster tempo). These studies of music processing in infancy will contribute to the ongoing debate about the biological and cultural foundations of music. References 1 Rothbaum, F. et al. (2000) Attachment and culture: security in the United States and Japan. Am. Psychol. 55, 1093–1104
327
2 Bowlby, J. (1969) Attachment and Loss: Vol 1. Attachment, Basic Books 3 Waters, E. and Cummings, E.M. (2000) A secure base from which to explore close relationships. Child Devel. 71, 164–172 4 Wallin, N. et al. (2000) The Origins of Music, MIT Press 5 Saffran, J.R. and Griepentrog, G.J. (2001) Absolute pitch in infant auditory learning: evidence of developmental reorganization. Dev. Psychol. 37, 74–85
Sandra E. Trehub Dept of Psychology, University of Toronto at Mississauga, Mississauga, Canada L5L 1C6. e-mail: [email protected]
Making connections about brain connectivity James V. Stone and Rolf Kötter The Functional Brain Connectivity workshop was organized by Rolf Kötter and Karl Friston, and held in Düsseldorf, Germany, on 4–6 April, 2002.
This three-day workshop on brain connectivity comprised 15 presentations by speakers using a wide variety different research techniques, including functional magnetic resonance imaging (fMRI), diffusion tensor imaging (DTI), optical imaging (OI) and computational modelling. Each speaker used only two transparencies, resulting in a very interactive and informal workshop. As might be expected, the resultant animated discussions between researchers from different scientific disciplines generated much heat, but also a few beams of much needed light. Functional, effective and structural connectivity
The very first presentation by Wim Vanduffel (Leuven Medical School, Belgium) led to a defining theme of the workshop: the distinction between functional, effective and structural connectivity. Essentially, structural connectivity refers to the direct anatomical connections between brain regions, whereas functional connectivity refers to a statistical dependency between the activities of different regions without regard to any underlying anatomical connection. These terms were coined during the pioneering attempts by Gerstein (University of Pennsylvania, Philadelphia, http://tics.trends.com
PA, USA), Aertsen (University of Freiburg, Freiburg, Germany) and colleagues to infer the nature of neuronal interactions from multi-electrode recordings, as recalled by Günther Palm (University of Ulm, Germany). Building on these early discussions, the distinction between functional and effective connectivity was first formally identified by Karl Friston (Wellcome Department of Cognitive Neurology, London, UK) in 1994. With Friston on hand to resolve disagreements over these definitions, it emerged that his original motivation for making such a distinction was to force researchers to be explicit about the nature of putative relationships among brain regions. Specifically, Friston argued that establishing effective connectivity between brain regions demands a precise mathematical specification of the nature of the proposed functional relationship in a way that functional connectivity does not. In other words, effective connectivity implies a precise mathematical model, whereas functional connectivity does not. The ensuing discussion soon showed that this seemingly clear distinction leads on to the question of what characterizes a model. Analyzing connectivity
Presentations of methods for analysing connectivity between brain regions were divided into two groups: those investigating structural connectivity using in vivo DTI (Rainer Goebel, University of Maastricht, The Netherlands; Martin A. Koch, Max Planck Institute of
Cognitive Neuroscience, Leipzig, Germany) and those using post-mortem human anatomical techniques (Karl Zilles, Research Centre Jülich, Germany). Owing to its technical limitations, DTI is presently no rival to in vivo tracer or post-mortem myeloarchitectonic methods. Indeed, Zilles emphasized that anatomical techniques imply that the Talairach atlas, which is the de facto standard for imaging coordinates, is essentially inaccurate. Zilles cogently made this point by presenting actual brain slices (which neatly subverted the conference limit of two transparencies). Despite the limitations of DTI, Goebel demonstrated the potential for this method to identify the anatomical pathways associated with activations observed using fMRI in the same subject. Methods for inferring connectivity included independent component analysis (James Stone, Sheffield University, UK), dynamic causal modelling (Karl Friston, Wellcome Functional Imaging Laboratory, London, UK), and structural equation modelling (Anthony Randy McIntosh, University of Toronto, Canada). A freeranging and provocative discussion on the role of causality in connectivity was also led by Ed Bullmore (University of Cambridge, UK). Presentations of methods for inferring the local functional topography of single cortical regions included optical imaging (Peter Buzás, Ruhr-Universität Bochum, Germany; Mark Hübener, Max-PlanckInstitut fur Neurobiologie, Martinsried,
1364-6613/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S1364-6613(02)01943-5