The Journal of Neuroscience, September 26, 2012 • 32(39):13339 –13342 • 13339

Mini-Review

Neural Mechanisms for the Abstraction and Use of Pitch

Information in Auditory Cortex
Xiaoqin Wang1,2* and Kerry M. M. Walker3*
1Tsinghua-Johns Hopkins Joint Center for Biomedical Engineering Research and Department of Biomedical Engineering, Tsinghua University, Beijing
100084, China, 2Department of Biomedical Engineering, Johns Hopkins University, Baltimore, Maryland 21025, and 3Department of Physiology, Anatomy
& Genetics, University of Oxford, Oxford, OX1 3PT, United Kingdom

Experiments in animals have provided an important complement to human studies of pitch perception by revealing how the activity of
individual neurons represents harmonic complex and periodic sounds. Such studies have shown that the acoustical parameters associ-
ated with pitch are represented by the spiking responses of neurons in A1 (primary auditory cortex) and various higher auditory cortical
fields. The responses of these neurons are also modulated by the timbre of sounds. In marmosets, a distinct region on the low-frequency
border of primary and non-primary auditory cortex may provide pitch tuning that generalizes across timbre classes.

Questions on pitch encoding mechanisms in auditory cortex Neurophysiological studies of pitch and periodicity coding in
Animal studies have provided useful information at the neuronal auditory cortex
level for understanding pitch processing mechanisms in auditory Following the discovery that macaque monkeys (Tomlinson and
cortex. Auditory cortex in non-human primates contains a Schwarz, 1988) and cats (Heffner and Whitfield, 1976) can per-
“core” region, composed of primary auditory cortex (A1), rostral ceive the pitch of missing fundamental harmonic complex
area (R), and rostral-temporal area (RT) (Fig. 1 A, B). The core is sounds, neurophysiologists embarked on attempts to identify
surrounded by “belt” and “para-belt” regions (Kaas and Hackett, neurons in A1 that have the same periodicity tuning for pure
2000; Hackett et al., 2001). Other mammals also have a “core” tones and missing fundamental sounds. Initial attempts per-
region in auditory cortex [A1 and anterior auditory field area formed in awake monkeys were unsuccessful. Two research
(AAF)], surrounded by secondary areas (ferret, Wallace et al., 1997; groups found that while the pure tone frequency tuning of many
Bizley et al., 2005; cat, Winer and Lee, 2007) (Fig. 1C,D). The simi- A1 neurons (i.e., characteristic frequency; CF) is consistent with
larities in auditory cortex organization make it possible to compare the neuron’s responses to harmonic tone complexes, their peri-
data obtained from different species, although one also must bear in odicity tuning does not extend to harmonic sounds with missing
mind the differences among species during this practice. fundamentals (Schwarz and Tomlinson, 1990; Fishman et al.,
There are several scenarios of how pitch could be encoded in 1998). Such results dispelled early expectations of finding a single
auditory cortex. Cortical areas with an orderly tonotopic organiza- neuron explanation of pitch perception at the level of A1.
tion, like A1, can represent pitch in frequency regions corresponding Several studies have, nevertheless, demonstrated that A1 neu-
to individual harmonic components, orthogonal to the tonotopic rons are tuned to a variety of acoustical cues that may be impor-
axis (Langer et al., 1997). Alternatively, pitch can be extracted by tant prerequisites to pitch encoding. Neurons in A1 are arranged
neurons tuned to low frequencies in a tonotopically organized cor- according to their CF, forming a tonotopic map of frequency
tical area (Bendor and Wang, 2005) or in a specialized cortical region preference across its surface. In the A1 of monkeys, the low-
(Griffiths and Hall, 2012). One would also like to know in any of frequency harmonics of periodic click trains are represented as
such scenarios whether a representation of pitch can be generalized distinct areas of activation across the low-CF region of the tono-
to all types of sounds bearing the same pitch (including missing topic map (Steinschneider et al., 1998). This might serve to rep-
fundamental sounds), and whether a particular neural representa- resent the resolved harmonics of periodic sounds (Oxenham,
tion is linked to pitch perception measured behaviorally. We will 2012). The temporal envelope modulations of higher (presum-
review below studies that address these questions. ably unresolved) harmonics, on the other hand, are represented
in the phase-locked responses of high-CF neurons (Steinsch-
neider et al., 1998). Therefore, acoustical properties that are
Received April 4, 2012; revised July 18, 2012; accepted July 23, 2012.
necessary for spectral- and temporal-based pitch extraction pro-
This work is supported by a grant from Tsinghua University (X.W.), NIH Grant R01 DC03180 (X.W.) and a Well-
come Trust Principal Research Fellowship (Andrew J. King). cesses are represented across the population of neurons in A1.
*X.W. and K.M.M.W. contributed equally to this work. A subpopulation of “multi-peaked” neurons have been iden-
Correspondence should be addressed to either of the following: Dr. Xiaoqin Wang, Department of Biomedical tified in cats (Sutter and Schreiner, 1991) and marmosets (Kadia
Engineering, Johns Hopkins University School of Medicine, 720 Rutland Avenue, Traylor 410, Baltimore, MD 21025, and Wang, 2003) which may be capable of harmonically fusing
E-mail: xiaoqin.wang@jhu.edu, or Dr. Kerry Walker, Department of Physiology, Anatomy & Genetics, University of
Oxford, Sherrington Building, Parks Road, Oxford, OX1 3PT, UK, E-mail: kerry.walker@dpag.ox.ac.uk.
complex sounds. These neurons respond preferentially not just
DOI:10.1523/JNEUROSCI.3814-12.2012 to a particular CF, but also to tones at frequencies that are har-
Copyright © 2012 the authors 0270-6474/12/3213339-04$15.00/0 monically related to their CF, particularly 1.5*CF and 2*CF. In
13340 • J. Neurosci., September 26, 2012 • 32(39):13339 –13342 Wang and Walker • Neural Mechanisms for Pitch

some cases, spiking responses to the CF in

the presence of its harmonics were en-
hanced over the response to CF alone (Ka-
dia and Wang, 2003). Similarly, a small
proportion (12%) of neurons in the two
core auditory cortical fields of awake ferrets
(A1 and AAF) have been shown to be har-
monically sensitive (Kalluri et al., 2008). In
these neurons, linear spectro-temporal fil-
ters computed from the cell’s response to
inharmonic sounds could not accurately
predict the response to harmonic complex
sounds. These studies show that multiple
harmonic components of complex periodic
sounds are integrated and represented as
spike rate codes in a subpopulation of neu-
rons in A1.
Extracellular recordings and intrinsic
optical imaging of responses in gerbil and
cat A1 to pure tones and sinusoidally
amplitude-modulated (SAM) tones have
suggested that a topographic gradient of
best modulation frequencies may exist for
complex, periodic sounds that is distinct
from, or even orthogonal to, the tonotopic
map (Schulze and Langner, 1997; Schulze et
al., 2002; Langner et al., 2009). Topographic
organization of best modulation rates has
not yet been found in primate A1 (Schwarz
and Tomlinson, 1990; Fishman et al., 1998).
In an intrinsic optical imaging study of pri-
mary and secondary auditory cortices in fer-
rets, Nelken et al. (2008) did find a gradient
of best modulation tuning for SAM tones,
but here the gradient ran approximately Figure 1. Schematics of auditory cortex across four species: A, Macaque; B, marmoset; C, cat; and D, ferret. In each panel, the top
parallel to the tonotopic map, in contrast to schematic shows an outline of the brain with auditory cortex indicted (dotted line, gray). The bottom schematic in each panel
earlier studies in humans (Langner et al., shows a closer view of the auditory cortex, with sulci (solid lines) and field boundaries (dotted lines). Core auditory cortex is shaded
gray, and the orientations of known tonotopic maps are indicated with an arrow from low (L) to high (H) frequencies. A1, Primary
1997) and gerbils (Schulze et al., 2002). auditory cortex; A2, secondary auditory area; AAF, anterior auditory field; ADF, anterior dorsal field; AL, anterolateral belt; CL,
Moreover, although periodicity gradients caudolateral belt; CM, caudomedial belt; DZ, dorsal zone; ML, mediolateral belt; MM, mediomedial belt; PAF, posterior auditory
were also observed for high-pass click trains field; PPF, posterior pseudosylvian field; PSF, posterior suprasylvian field; r, rostral field; RM, rostromedial belt; RT, rostral temporal
and high-pass iterated rippled noise, there field; RTL, rostrotemporal lateral belt; RTM, rostrotemporal medial belt; VP, ventral posterior field.
was no consistent arrangement of periodic-
ity preference gradients across the three neurons, but not single neurons, provided sufficient F0 discrim-
stimulus types (Nelken et al., 2008). The stimulus specificity of these ination to account for ferrets’ thresholds on an equivalent behav-
periodicity preferences means that they cannot be interpreted as ioral task, in which ferrets were trained to categorize the same
generalized pitch maps. Neuronal sensitivity to confounded fea- sounds as “low” or “high.” This group further showed that al-
tures, such as harmonic spacing due to critical bandwidths (Fishman though neurons that were sensitive to the periodicity of artificial
et al., 2000) or cochlear distortion products (Wiegrebe and Patter- vowels could be found across 5 examined fields of primary and
son, 1999), have been offered as potential explanations for the ob- secondary auditory cortex, these neurons were not selective (i.e.,
served organization of modulation preferences for SAM tones.
specialized) for pitch (Bizley et al., 2009). A neuron that was sensitive
Although A1 neurons can extract acoustical cues that are nec-
to vowel pitch almost always carried information about the timbre
essary to compute the pitch of a variety of periodic sounds, the
or spatial location of the vowel as well. For a subset of these neurons,
response properties of a single A1 neuron would be insufficient to
represent the pitch of the entire range of pitch-evoking sounds, feature representations were multiplexed within separate response
particularly those with missing fundamentals. For most sounds time windows, so the pitch and timbre of vowels can be invariantly
that humans and animals encounter in their natural environ- represented in a single auditory cortical neuron (Walker et al., 2011).
ments, such as vocal calls, these neural representations may be It may therefore be instructive for future studies to examine pitch
sufficient for pitch extraction. This possibility was examined by tuning in multiple time windows throughout a neuron’s response to
Bizley et al. (2010), who trained statistical “neurometric” algo- sounds, since tuning properties across the onset, sustained, and off-
rithms to discriminate the periodicity of artificial vowel sounds set windows can be fundamentally different (Wang et al. 2005).
(i.e., bandpass-filtered click trains) based on the responses of Given that humans and animals do experience a percept of
neurons in ferret auditory cortex. They found that neurometrics pitch that generalizes across a variety of sounds with the same
based on the response of small populations of auditory cortical periodicity (including missing fundamental sounds), it seems
Wang and Walker • Neural Mechanisms for Pitch J. Neurosci., September 26, 2012 • 32(39):13339 –13342 • 13341

reasonable to expect to find neurons at some level of the auditory Computationally, it is important to differentiate between
system that integrate the periodicity cues described above to neurons (or cortical regions) that extract pitch embedded in
compute a stimulus-invariant pitch representation. Such neu- complex sounds (such as harmonic complex) and those that
rons may be distributed throughout the auditory cortex, but hu- bear pitch information. This requires examining whether
man imaging studies suggest that pitch neurons are likely to be pitch is specifically or uniquely represented by the neuron or
concentrated in a region of auditory cortex that is specialized to cortical region under study, as well as determining that the
encode the periodicity of temporally regular sounds (Griffiths physiological signal of the neurons corresponds with the ani-
and Hall, 2012). Bendor and Wang (2005) have provided evi- mal’s perception of pitch. Neural responses bearing pitch in-
dence for pitch extraction by single neurons in the auditory cor- formation or encoding acoustic parameters associated with
tex of awake marmosets. “Pitch-selective neurons” described in pitch can be found throughout much of the ascending audi-
this study were defined as those whose CF for pure tones matched tory pathway (Cariani and Delgutte, 1996a,b), though their
their periodicity tuning for missing fundamental harmonic com- specificity for pitch may increase at successive higher process-
plex sounds, where the spectral components of the latter all lay ing stages. Neural representations earlier in the system could
outside of the neuron’s excitatory-frequency response area. The serve as precursors to the neurons that ultimately compute
region containing the pitch-selective neurons is confined to the pitch, but they may not represent the final stages of pitch
low-frequency border of A1, R, and lateral belt areas. Approxi- processing. Technically, it is crucial to apply rigorous controls
mately 39% of neurons within this anterolateral pitch region to rule out influences by such factors as acoustic artifacts and
were classified as pitch-selective neurons based on multiple cochlear distortions before a neuron or cortical region is con-
criteria. Using temporally jittered click trains and iterated rip- sidered “pitch-selective.”
pled noises, Bendor and Wang (2010) further demonstrated
that these pitch-selective neurons were sensitive to the tempo-
ral regularity of sounds, unlike modulation-sensitive neurons References
outside the pitch area, which are instead tuned to repetition Bendor D, Wang X (2005) The neuronal representation of pitch in primate
auditory cortex. Nature 436:1161–1165. CrossRef Medline
rate regardless of the waveform’s temporal regularity. A recent Bendor D, Wang X (2006) Cortical representations of pitch in monkeys and
study has reported evidence of pitch perception by marmosets humans. Curr Opin Neurobiol 16:391–399. CrossRef Medline
(Osmanski et al., 2011). Bendor D, Wang X (2010) Neural coding of periodicity in marmoset audi-
A cortical region analogous to the pitch region reported by tory cortex. J Neurophysiol 103:1809 –1822. CrossRef Medline
Bendor and Wang (2005) has been identified in several human Bizley JK, Nodal FR, Nelken I, King AJ (2005) Functional organization of
imaging studies, but has yet to be identified in other animal spe- ferret auditory cortex. Cereb Cortex 15:1637–1653. CrossRef Medline
cies for which behavioral evidence of pitch perception is estab- Bizley JK, Walker KM, Silverman BW, King AJ, Schnupp JW (2009) Inter-
dependent encoding of pith, timbre and spatial location in auditory cor-
lished. But if it does exist, one shall expect to find it in similar
tex. J Neurosci 29:2064 –2075. CrossRef Medline
low-frequency borders of primary and non-primary cortical Bizley JK, Walker KM, King AJ, Schnupp JW (2010) Neural ensemble codes
fields. While such a region was not specifically investigated in for stimulus periodicity in auditory cortex. J Neurosci 30:5078 –5091.
the study by Bizley et al. (2009), that study observed an over- CrossRef Medline
representation of neurons sensitive to the pitch of artificial Cariani PA, Delgutte B (1996a) Neural correlates of the pitch of complex
vowels in the low-frequency borders of A1 and tonotopic sec- tones. I. Pitch and pitch salience. J Neurophysiol 76:1698 –1716. Medline
ondary fields in the ferret. Ultimately, lesion or cortical inac- Cariani PA, Delgutte B (1996b) Neural correlates of the pitch of complex
tones. II. Pitch shift, pitch ambiguity, phase invariance, pitch circularity,
tivation studies will be a necessary complement to these
rate pitch, and the dominance region for pitch. J Neurophysiol 76:1717–
single-neuron investigations to establish the behavioral rele- 1734. Medline
vance of putative pitch codes. Fishman YI, Reser DH, Arezzo JC, Steinschneider M (1998) Pitch vs. spec-
tral encoding of harmonic complex tones in primary auditory cortex of
Outstanding questions and directions for future work the awake monkey. Brain Res 786:18 –30. CrossRef Medline
There are several key questions that remain to be addressed in Fishman YI, Reser DH, Arezzo JC, Steinschneider M (2000) Complex tone
future studies of pitch encoding in auditory cortex. First, is there processing in primary auditory cortex of the awake monkey. II. Pitch
versus critical band representation. J Acoust Soc Am 108:247–262.
more than one “pitch center” in auditory cortex? If so, what are
CrossRef Medline
specific roles played by these different pitch-processing centers? Griffiths TD, Hall DA (2012) Mapping pitch representation in neural en-
To answer these questions, researchers need to pursue investiga- sembles with fMRI. J Neurosci 32:13343–13347. CrossRef
tions to identify cortical regions, in particular in the secondary Hackett TA, Preuss TM, Kaas JH (2001) Architectonic identification of the
auditory cortex, that extract pitch embedded in complex sounds core region in auditory cortex of macaques, chimpanzees, and humans.
using a wide range of stimuli. Second, what are underlying mech- J Comp Neurol 441:197–222. CrossRef Medline
anisms that allow any “pitch neuron” to perform pitch extraction Heffner HH, Whitfield IC (1976) Perception of the missing fundamental by
cats. J Acoust Soc Am 51:520 –529.
computations? To understand such questions, one has to employ
Kaas JH, Hackett TA (2000) Subdivisions of auditory cortex and processing
techniques beyond extracellular recordings, such as intracellular streams in primates. Proc Natl Acad Sci U S A 97:11793–11799. CrossRef
and two-photon optical recordings. Third, is a putative pitch Medline
region involved in pitch perception? While a demonstration of Kadia SC, Wang X (2003) Spectral integration in A1 of awake primates:
parallel properties between an animal’s behavioral responses and neurons with single and multipeaked tuning characteristics. J Neuro-
corresponding neural responses is a useful step toward answering physiol 89:1603–1622. Medline
such a question, a more convincing demonstration would be to Kalluri S, Depireux DA, Shamma SA (2008) Perception and cortical neural
coding of harmonic fusion in ferrets. J Acoust Soc Am 123:2701–2716.
show that the interruption of neural activity in a putative pitch
CrossRef Medline
region will lead to alterations in an animal’s pitch perception Langner G, Dinse HR, Godde B (2009) A map of periodicity orthogonal to
performance. Newly emerged optogenic techniques could be a frequency representation in the cat auditory cortex. Front Integr Neurosci
useful tool in this line of research in addition to other inactivation 3:27. Medline
methods. Nelken I, Bizley JK, Nodal FR, Ahmed B, King AJ, Schnupp JW (2008) Re-
13342 • J. Neurosci., September 26, 2012 • 32(39):13339 –13342 Wang and Walker • Neural Mechanisms for Pitch

sponses of auditory cortex to complex stimuli: functional organization Click train encoding in primary auditory cortex of the awake monkey:
revealed using intrinsic optical signals. J Neurophysiol 99:1928 –1941. evidence for two mechanisms subserving pitch perception. J Acoust Soc
CrossRef Medline Am 104:2935–2955. CrossRef Medline
Osmanski MS, Bendor DA, Wang X (2011) Dual mechanisms for process- Sutter ML, Schreiner CE (1991) Physiology and topography of neurons with
ing pitch in marmosets (Callithrix jacchus). Soc Neurosci Abstr 38:691.08. multipeaked tuning curves in cat primary auditory cortex. J Neurophysiol
Oxenham AJ (2012) Pitch perception. J Neurosci 32:13335–13338. 65:1207–1226. Medline
CrossRef Tomlinson RW, Schwarz DW (1988) Perception of the missing fundamen-
Schulze H, Langner G (1997) Periodicity coding in the primary auditory tal in nonhuman primates. J Acoust Soc Am 84:560 –565. CrossRef
cortex of the Mongolian gerbil (Meriones unguiculatus): two different Medline
coding strategies for pitch and rhythm? J Comp Physiol A 181:651– 663. Walker KM, Bizley JK, King AJ, Schnupp JW (2011) Multiplexed and robust
CrossRef Medline representations of sound features in auditory cortex. J Neurosci 31:
Schulze H, Hess A, Ohl FW, Scheich H (2002) Superposition of horseshoe- 14565–14576. CrossRef Medline
like periodicity and linear tonotopic maps in auditory cortex of the Mon- Wallace MN, Roeda D, Harper MS (1997) Deoxyglucose uptake in the ferret
golian gerbil. Eur J Neurosci 15:1077–1084. CrossRef Medline auditory cortex. Exp Brain Res 117:488 –500. CrossRef Medline
Schwarz DW, Tomlinson RW (1990) Spectral response patterns of auditory Wiegrebe L, Patterson RD (1999) Quantifying the distortion products gen-
cortex neurons to harmonic complex tones in alert monkey (Macaca erated by SAM noise. J Acoust Soc Am 106:2709 –2718. CrossRef Medline
mulatta). J Neurophysiol 64:282–298. Medline Winer JA, Lee CC (2007) The distributed auditory cortex. Hear Res 229:3–13.
Steinschneider M, Reser DH, Fishman YI, Schroeder CE, Arezzo JC (1998) CrossRef Medline