🧠 Functional Anatomy of Musical Processing in Absolute Pitch

📋 Study Overview

🎯 Research Question

Which brain regions differ between musicians with and without absolute pitch (a) while they simply listen to pairs of tones, and (b) while they classify the interval as major or minor?

No pitch-naming task was performed inside the scanner. Note naming happened only in the screening that assigned participants to groups — which matters for reading everything below.

Following Schlaug's 1995 report of structural brain differences (planum temporale asymmetry), Zatorre's team used functional imaging to see which regions change blood flow during tonal processing. The paper itself claims no priority, and records that electrophysiological work on AP (P300 studies) already existed.

🔬 Methodology

Participants (N = 20)

• AP group (n=10): Right-handed musicians who identified 100 synthetic tones with average error ≤0.6 semitones (mean error: 0.16 semits). Two additional self-identified AP musicians were rejected for failing this screen
• RP group (n=10): Right-handed musicians confirmed to have good relative pitch but no AP (mean screening error: 2.44 semits)
• Both groups matched for age (~25 years); AP group had more musical experience (18.2 vs 13.1 years)

Brain Imaging Protocol

• Technology: PET (Scanditronix PC-2048B, 15-slice) measuring cerebral blood flow (CBF) via O-15 water bolus. MRI (1.5T Phillips) for anatomy and PT morphometry
• Stimuli: Sawtooth waves (15 harmonics), F#3 to C#5, 500ms each, presented in pairs forming ascending/descending minor or major thirds
• Three conditions:
  • Noise (baseline): Pairs of noise bursts acoustically matched to tones — press key after each pair
  • Tones: Listen to tone pairs forming musical intervals — just listen and press key (no labeling required)
  • Minor/Major: Same tone pairs — classify each interval as minor or major (“a judgment that may be accomplished via RP” — the paper argues AP possessors likely did not need it)
• Key design insight: By comparing "just listening" vs. "active classification," the study separates passive pitch perception from active labeling processes

How the group differences were assessed

Each group's map was computed separately (active condition minus noise), thresholded at t = 3.5 (P < 0.0004, uncorrected for multiple comparisons; ~0.58 false positives expected per search volume of 200 resolution elements). No direct AP vs. RP statistical contrast is reported in the paper. The differences described below are differences between two independently thresholded maps, with n=10 per group. The correct reading is "it appeared in one map and not in the other" — not "the groups differ significantly."

📊 Key Findings

1. Left DLF Cortex — AP Activates It Just by Listening

Tones minus Noise: AP musicians showed strong activation of left posterior dorsolateral frontal cortex (Brodmann area 8/6, t=4.45) while simply listening to tones. In the RP group, no trace of blood-flow change appeared above threshold in this region — which, as noted above, is not the same as a significant group difference.

• This region is associated with conditional associative learning — mapping stimuli to labels (Petrides, 1990)
• AP subjects reported being "generally aware of the correct note names" even though they were only asked to listen
• The authors' reading: in AP possessors such associations occur spontaneously and form the basis of AP labeling. The paper offers this as what the activation "may reflect" — the evidence is one PET focus plus the participants' own debriefing report, not a demonstration

2. The Twist — RP Musicians Activate the Same Region When Labeling

Minor/Major minus Noise: When RP musicians had to classify intervals (a labeling task), they also activated left DLF cortex — in nearly the same location as AP subjects.

• This means left DLF is not unique to AP — the authors propose that the posterior DLF plays a key role in a distributed network for verbal–tonal associations
• AP subjects activate it spontaneously; RP subjects activate it only when a task demands labeling
• The difference is not what brain region is used, but when it is engaged

3. A Dissociation in Right Inferior Frontal Cortex — and the Reading the Authors Propose

In the tones condition, right inferior frontal cortex was active in both groups (AP t=3.69; RP t=4.07). What stands out is that in the Minor/Major task it disappears in the AP group only — a loss of activation, not a region AP possessors never use:

The hypothesis, in the authors' words. They present it explicitly as "one possible interpretation": the right inferior frontal region, present only in the RP group in the minor/major condition, "may reflect maintenance of pitch information in auditory tonal working memory." If that is right, AP possessors "may not need access to this mechanism" for interval classification, because they name each note instead. This is a hypothesis, not a measurement — the study measured blood flow, not working memory, and an absence of suprathreshold activation in a group of 10 is not evidence that the process is absent.

4. Auditory Cortex — Nearly Identical Foci, but the Door Is Not Closed

In the tones-minus-noise comparison, both groups showed nearly identical loci of blood-flow increase in superior temporal gyrus bilaterally (Brodmann areas 22/42). That led the authors to conclude that AP "would not appear to involve differences at the level of the initial stages of perceptual analysis."

They do not close the perceptual door, though. The paper ends by proposing that AP "may result from an interaction between computations in the superior temporal area and the engagement of a network of brain regions, particularly the posterior DLF cortex," and grants that it is "possible that AP arises from some qualitatively different neural process within the superior temporal region" — indexed, they suggest, by the planum temporale morphology below.

5. PT Morphometry — Nuanced Results

• Left PT was larger in the AP group (4950 mm³) than in a reference sample of 50 right-handers unselected for musical training (4238 mm³; Mann–Whitney U=131, P<0.03). The RP group's PT did not differ from that sample in either hemisphere
• BUT: No significant difference between AP and RP groups (4950 vs 4160, large variability)
• No exaggerated asymmetry (unlike Schlaug 1995) — the right PT was also larger in the AP group, albeit not significantly so, so the paper concludes it did not find exaggerated asymmetry "per se"
• Correlation found: left PT volume correlated with pitch-naming error across ALL subjects (r=−0.39, P=0.05) — larger PT, lower error

6. Behavioral Data

AP subjects were significantly more accurate on the Minor/Major task (96.7% vs 82.8%, P<0.01) but not faster (2224 vs 2335 ms, ns). AP helps accuracy, but doesn't speed up relative pitch tasks.

💡 Main Conclusions

"The findings of the present study suggest that no one regional activation pattern is unique to AP. Rather, the areas recruited depend upon the task demands, and the availability of specific processing mechanisms." — Zatorre et al., 1998 (p. 3177)

Key Implications:

• AP is not a unique brain circuit: it recruits a labeling network (left DLF) that non-AP musicians also use — but AP possessors engage it spontaneously, without being asked to label. In the authors' words, "this network can also be used by musicians without AP… whenever associations can be made to the relation between pitches rather than to a single pitch"
• AP does not look like "better hearing": superior temporal foci were nearly identical across groups, which led the authors to write that AP "would not appear to involve differences at the level of the initial stages of perceptual analysis." They stop short of ruling perception out, and end by proposing an interaction between superior temporal computations and the frontal network
• A possible working-memory shortcut — offered as a hypothesis: right inferior frontal activity, present in RP but not AP during interval classification, "may reflect" holding pitch in auditory working memory; if so, AP possessors "may not need access to this mechanism." What the study actually measured points to accuracy, not speed: AP subjects were more accurate (96.7% vs 82.8%) but no faster (2224 vs 2335 ms, not significant)
• Associative retrieval: the paper frames the left DLF finding in terms of conditional associative learning in Petrides' sense — arbitrary stimulus–response associations (Petrides, 1990). It does not describe AP as "overlearned," and draws no analogy to reading or colour naming
• Structure–performance correlation, with the paper's own caveats: left PT volume correlated with pitch-naming error across all 20 subjects (r=−0.39, P=0.05, at the edge of significance). The authors call it “direct evidence for the possible existence of a structure–function relationship" and ask for caution on three stated grounds: (i) the AP group's PT differs from the reference sample but not from the RP group; (ii) there was no evidence of any blood-flow change in or near the PT in either task — the structure correlates, the function did not appear; (iii) the result differs in form from Schlaug 1995, which reported a difference in asymmetry rather than in absolute left volume. Their closing line: "the precise role played by this region remains to be determined"

⚠️ Limitations & Context

Study Limitations

• No direct group contrast, and an uncorrected threshold: maps were thresholded separately at t=3.5 (P<0.0004, uncorrected), n=10 per group, with no AP × RP statistical test reported
• The paper's three caveats on the planum temporale: the asymmetry does not distinguish AP from RP; no blood-flow change appeared near the PT in either task; and the result differs in form from Schlaug 1995
• A null the authors flag themselves: "no significant differences in PT volume between AP and RP groups, possibly because of the small sample size and large variability"
• Interpretations marked as possibilities: the working-memory reading, the associative mechanism and the temporal-lobe origin are all given as may / one possible interpretation, never as findings
• Not measured here: no pitch naming inside the scanner; and the AP group had ~5 more years of musical training (18.2 vs 13.1), which the paper neither controls nor discusses
• Cross-sectional: cannot show how these patterns develop, or whether an adult learner would produce them

Historical Context (1998 vs 2020s)

🔗 Related Research

• Structural basis: Schlaug et al. (1995) — stronger leftward planum temporale asymmetry in musicians with AP (in that sample, mostly from a smaller right PT). This study reports a difference in left PT volume instead, which is a different form of result
• Connectivity: Loui et al. (2011) — hyperconnectivity in bilateral superior temporal structures, with left STG–MTG tract volume predicting AP performance. (The auditory–frontal link via the arcuate fasciculus is a different paper — Loui, Alsop & Schlaug 2009 — and it is about musical disorders)
• 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; neural mechanisms in learners unknown

📖 Access Full Study

📚 Full Citation