NEUROIMAGING 1995

🧠 Structural Brain Asymmetry in Musicians with Absolute Pitch

⚠️ Historical study (1995): Published in Science, an early report of a structural brain difference associated with absolute pitch. Its core finding — that AP musicians have exaggerated leftward planum temporale asymmetry — has had a mixed record since. Zatorre et al. (1998), also in this collection, report that they “did not find an exaggerated asymmetry per se.” Note also that relative to the other musicians, the AP group’s asymmetry comes mainly from a smaller right planum temporale rather than a larger left one (1097 vs 1043 on the left; 611 vs 830 on the right). Against the non-musicians the arithmetic flips: the left PT is 201 mm² larger and the right 125 mm² smaller. The study could not determine causality: does the brain difference enable AP, or does AP training reshape the brain? The paper itself acknowledges this open question. Recent adult training research (Wong et al. 2025) adds urgency to resolving it.

📋 Study Overview

Title:

In vivo evidence of structural brain asymmetry in musicians

Authors:

Gottfried Schlaug, Lutz Jäncke, Yanxiong Huang, Helmuth Steinmetz

Published:

Science, February 3, 1995; 267(5198):699-701


🎯 Research Question

Are there structural brain differences that distinguish musicians with absolute pitch from those without?

The paper's own stated question was broader: whether professional musicians as a group show different planum temporale asymmetry from non-musicians. The AP/non-AP split was one of two planned orthogonal contrasts within the 30 musicians — and it turned out to carry most of the group effect. The paper is titled "In vivo evidence of structural brain asymmetry in musicians," and its abstract concludes about "outstanding musical ability," not about AP.

Prior to this study, absolute pitch was understood behaviorally and psychologically, but no one had looked inside the living brain to see if AP possessors showed distinct anatomical features. Using MRI technology (cutting-edge in 1995), Schlaug's team investigated the planum temporale — an auditory brain region known to show asymmetry in most people.


🔬 Methodology

Participants (N = 60)

  • 30 professional musicians (right-handed), recruited via announcements at three German music schools and through the authors' personal contacts:
    • 11 with AP — mean age 27 (SD 5)
    • 19 without AP — mean age 26 (SD 4)
    • All 30 played keyboard; 14 of them preferred string instruments, the other 16 were keyboard players only. In a study about absolute pitch this is not decoration — the entire sample trained on a fixed-pitch, labelled instrument
  • 30 non-musicians (control) — mean age 26 (SD 3), matched for age, sex, and handedness
  • AP verification — none formal. In the authors' words: "no formal test for perfect pitch was applied." Participants were asked whether they could sing a named tone and name a heard tone without a reference; the 11 who said yes had the claim checked against music-school examination records for a pitch discrimination task — distinguishing whether two tones differ, which is a different ability from identifying a tone, and one most professional musicians have. AP status in this study is self-report. Since the AP/non-AP split is the axis the whole page turns on, this is the study's most serious methodological limitation

Brain Imaging Protocol

  • Technology: Siemens 1.5T MRI — 128 contiguous sagittal slices (1.17 mm thick), voxel size 1.00 × 1.00 × 1.17 mm
  • Region of interest: Planum temporale (PT) — auditory association cortex on the superior surface of the temporal lobe, involved in pitch and language processing
  • Measurement: Curved length of PT on each slice × slice thickness = cortical surface area (mm²) for left and right hemispheres
  • Asymmetry index: δPT = (R − L) / [0.5(R + L)] — negative values indicate leftward asymmetry
  • Blinded analysis: Two independent observers, blinded to musician/non-musician status and hemisphere side. Inter-observer correlation: r = 0.93

📊 Key Findings

1. Exaggerated Leftward Asymmetry — Only in AP Musicians

Group N δPT (SD) Left PT (mm²) Right PT (mm²)
Musicians with AP 11 −0.57 (0.21) 1097 (202) 611 (105)
Musicians without AP 19 −0.23 (0.17) 1043 (183) 830 (178)
All musicians (the paper's own headline row) 30 −0.36 (0.25) 1063 (189) 750 (187)
Non-musicians 30 −0.23 (0.24) 896 (236) 736 (263)

AP musicians vs. musicians without AP: F(1,57) = 16.18, p < 0.001. Musicians (all) vs. non-musicians: F(1,57) = 5.12, p = 0.028. Both tests were run on δPT only — raw PT size was never tested statistically in this paper; the size columns above are descriptive means.

Note where the asymmetry comes from. Between AP and non-AP musicians, the left PT differs by about 5% (1097 vs 1043, inside one standard deviation) while the right PT differs by about 26% (611 vs 830). In this dataset the exaggerated leftward asymmetry of the AP group is driven mainly by a smaller right PT, not a larger left one. The paper does not comment on this.

2. Non-AP Musicians Did Not Differ From Non-Musicians

Musicians without AP showed essentially the same PT asymmetry (δPT = −0.23) as non-musicians (δPT = −0.23). Years of professional-level practice, on their own, did not produce the exaggerated asymmetry.

But the group effect is real, and it is the paper's title. Musicians as a whole did differ from non-musicians in asymmetry — δPT −0.36 vs −0.23, F(1,57) = 5.12, p = 0.028 — and the abstract closes on it: "the results indicate that outstanding musical ability is associated with increased leftward asymmetry of cortex subserving music-related functions." What the AP split shows is where that group effect came from. In the authors' words, "possession of perfect pitch explained most of the variation in the degree of PT asymmetry among musicians (P < 0.001)" — most, not all.

3. What Is the Planum Temporale?

The PT is a triangular region on the upper surface of the temporal lobe, part of the auditory association cortex. It is notable because:

  • It overlaps with Wernicke's area (language comprehension)
  • Its leftward asymmetry appears in the human fetus between the 29th and 31st gestational week, which the authors read as evidence that "prenatal factors are likely to play a role"
  • The paper cites PET evidence that this region is involved in music perception (Mazziotta et al., 1982, a tonal memory task, where listeners who reported "specific, highly organized" analytic strategies showed left-greater-than-right temporoparietal activation). The paper does not describe the region as doing "categorical" processing — that word does not appear in it
  • Left-handers tend to have more symmetrical PT, which correlates with atypical lateralization of both language and music

4. Nature vs. Nurture — The Unresolved Question

The paper explicitly acknowledges the causality dilemma:

  • Prenatal factors likely play a role — PT asymmetry is present before birth
  • But: maturation of fiber tracts and intracortical neuropil is "still progressing by the age of seven" — the paper gives no endpoint — so "it remains uncertain whether gross anatomy may also be susceptible to some postnatal plastic change"
  • This study collected no age-of-training data. The claim comes from Sergeant (1969), and it runs the other way round: "almost all musicians who began training before the age of seven had perfect pitch, but almost none of those beginning after the age of 11." That is P(AP | early start), not P(early start | AP) — a much stronger statement, and not one this paper tested
  • And it was a premise, not a result: the authors write that "in distinguishing between musicians with or without perfect pitch, we were guided by" that same Sergeant study. The operational criterion they describe, though, is another one (note 19): self-report confirmed against school records. The paper did not collect age of onset, so it cannot serve as independent evidence for a critical period
  • Either way, the anatomy is consistent with both explanations: a prenatal predisposition activated by early training, or early training reshaping still-developing anatomy

💡 Main Conclusions

"Our study demonstrates that individual variability in cognitive performance can covary with features of external brain morphology." — Schlaug et al., 1995 (p. 700)

Key Implications:

  • An anatomical correlate: the exaggerated leftward PT asymmetry tracks AP rather than musical training as such — musicians without AP looked like non-musicians. The authors put it as "most of the variation," not all of it
  • A language-adjacent region: the paper notes that the left PT coincides with Wernicke's speech area as identified by lesion studies. The authors themselves draw a musical, not a linguistic, conclusion — the finding "may suggest that the functional capacity of cortex shown to subserve musical functions increases with leftward structural asymmetry of this neural system." Reading AP as categorical labelling is a later interpretation, not one this paper makes
  • What was and was not tested: the ANOVA ran on δPT alone. The raw size columns are descriptive, so the page makes no claim that training "enlarges" the PT — and the right-hemisphere means of musicians and non-musicians (750 vs 736 mm²) do not support a bilateral enlargement anyway
  • Critical period — invoked here, not tested here: this study collected no age-of-training data at all. The authors cite Sergeant (1969), where "almost all musicians who began training before the age of seven had perfect pitch, but almost none of those beginning after the age of 11," and state they used that literature to guide their own AP/non-AP classification. That makes it a premise of the design rather than a finding of it
  • Foundation for neuroimaging: This 3-page paper launched decades of brain imaging research on AP, including DTI (Loui et al. 2011), fMRI, and PET studies

⚠️ Limitations & Context

Study Limitations

  • Correlation, not causation: Cannot determine if larger PT asymmetry predisposes to AP or results from AP-related training/experience. The paper acknowledges this explicitly
  • No AP test at all: "no formal test for perfect pitch was applied." AP status rests on self-report, checked against school records for a pitch discrimination task — a different ability. Compare Miyazaki 1988, which administered a scored identification test
  • Small AP group: N=11 AP musicians limits statistical power, though the effect size was large enough to reach p < 0.001
  • Handedness caveat: all participants were right-handed by design, but 6 of 60 (3 musicians, 3 non-musicians) were classified as non-consistent right-handers on Annett's 12-item questionnaire. Results may not generalize to left-handers, who tend to have a more symmetrical PT. Sex composition is not reported for the musician group; the non-musicians were sex-matched to them
  • Cross-sectional design: No longitudinal data showing whether PT asymmetry changes with training over time

Historical Context (1995 vs. 2020s)

📖 What Has Changed Since 1995:
This paper was widely interpreted as evidence that AP is "hardwired" — a structural brain feature you either have or don't. But the paper itself is more cautious, acknowledging that postnatal plasticity cannot be ruled out. Thirty years later, the causality question remains open. Loui et al. (2011) found enhanced white matter connectivity (not just volume) in AP brains. Wong et al. (2025) showed adults can develop functional AP with training — raising the tantalizing question: would their PT asymmetry change too? No longitudinal MRI study of adult AP learners has been published yet. This remains one of the great unanswered questions in the field.

🔗 Related Research

  • Follow-up structural imaging: Loui et al. (2011) - white matter connectivity (DTI) shows enhanced tracts in AP musicians
  • Functional imaging: Zatorre et al. (1998) - PET scan showing activation differences during pitch tasks
  • Follow-up: Keenan et al. (2001) — also examined planum temporale in AP possessors. Results across studies are mixed: Zatorre et al. (1998) did not replicate the exaggerated asymmetry
  • Adult trainability: Wong et al. (2025) - adults raised pitch-naming accuracy from 13.9% to 31.7% (chance 8.3%) after ~21 h of training, with 2 of 12 reaching AP-level performance; neuroplasticity question remains open

📖 Access Full Study


📚 Full Citation

Schlaug, G., Jäncke, L., Huang, Y., & Steinmetz, H. (1995). In vivo evidence of structural brain asymmetry in musicians. Science, 267(5198), 699–701. https://doi.org/10.1126/science.7839149