DEVELOPMENTAL NEUROSCIENCE Cortical indices of sound localization mature monotonically in early infancy Christopher Slugocki and Laurel J. Trainor Department of Psychology, Neuroscience & Behaviour, McMaster University, 1280 Main Street West, Hamilton, ON L8S4L8, Canada Keywords: auditory-evoked potentials, azimuth, development, mismatch negativity Abstract In human neonates, orienting behavior in response to an off-midline sound source disappears around the first postnatal month, only to re-emerge at ~4 months. To date, it is unclear whether sound localization processes continue to operate between postna- tal months 1 and 3. Here, we used an event-related potential, reflecting change detection in the auditory cortices, to measure the cortical responses elicited by large (Æ 90° relative to midline), infrequent changes in sound source location in 2-, 5-, 8- and 13- month-old infants. Both fast-negative mismatch negativity (MMN) N€ a€ at€ anen et al. (2007) and slow-positive mismatch response (MMR) Trainor et al. (2003) were elicited from all age groups. However, both components were smaller and the fast-negative component occurred later in the 2-month-old group than in older age groups. Additionally, the slow-positive component tended to diminish in amplitude with increasing age, whereas the fast-negative component grew larger and tended to occur earlier with increasing age. These results suggest that the cortical representation of sound location matures similarly to representations of pitch and duration. A subsequent investigation of 2-month-old infants confirmed that the observed MMR and MMN were elicited by changes in sound source location, and were not merely attributable to changes in loudness cues. The presence of both MMR and MMN in the 2-month-old group indicates that the cortex is able to detect changes in sound location despite the behavioral insensitivity observed around 1–3 months of age. Introduction Research suggests that the postnatal development of human spatial hearing is best approximated by a U-shaped function [reviewed by Muir et al. (1989)]. Neonatal orienting responses, as observed within the first postnatal month (Wertheimer, 1961; Crassini & Broerse, 1980), disappear between 1 and 3 months of age, only to re-emerge between 3 and 4 months of age, with decreased latency, greater sensitivity, and increasing absolute precision (Muir et al., 1979, 1989; Clifton et al., 1981; Muir & Hains, 2004). The domi- nant explanation for this dip in orienting behavior posits that neona- tal localization is driven by a subcortically mediated reflex that becomes suppressed at ~1 month of age. Suppression is thought to have a facilitative role in the transition to a more volitional and cor- tically mediated sound localization mechanism (Muir & Clifton, 1985; Muir & Hains, 2004). To date, it is not clear whether infants aged 1–3 months are actu- ally able to localize sounds and, if they are, whether they are doing so at a cortical level. Here, we addressed this question by measuring electroencephalographic indices of pre-attentive change detection in the auditory cortices, known as mismatch response (MMR) and mis- match negativity (MMN), in a cross-sectional sample of 2– 13-month-old infants. Developmental studies have consistently indi- cated a morphological transition in the mismatch component, with a slow-positive MMR being predominant in younger infants and diminishing in amplitude with age, and an adult-like fast-negative wave, resembling the MMN response of adults and older children, emerging in older infants and increasing with age [for reviews, see Trainor (2007, 2012a,b)]. In adults, the MMN originates primarily in the auditory cortex (for a review, see N€ a€ at € anen et al., 2007), and its morphological features, such as amplitude and latency, are highly correlated with behavioral measures of discrimination thresholds for many auditory features, including pitch (e.g. Novitski et al., 2004), gap detection (Trainor et al., 2001), and, of relevance to the current study, sound localization (Paavilainen et al., 1989; Schr€ oger, 1996; Schr€ oger & Wolff, 1996; Kaiser et al., 2000a,b; Altman et al., 2005, 2009; Deouell et al., 2006; Sonnadara et al., 2006; Pakarinen et al., 2007; R€ ottger et al., 2007; Spierer et al., 2007; Altmann et al., 2009; Vasilenko & Shestopalova, 2010; Grimm et al., 2012; Bennemann et al., 2013). The first experiment in the present study employed an oddball par- adigm, such that, in 80% of trials, a sound was presented from directly in front of the infant at midline, and in the remaining trials from a location Æ 90° relative to the midline. If infants are indeed localizing during the period of behavioral silence, we expect deviant trials to elicit one or both mismatch components from all age groups. Furthermore, if 2-month-old localization is driven primarily by sub- cortical processes, we expect the immature slow-positive MMR to dominate the 2-month-old response to deviant trials and an adult-like Correspondence: Laurel Trainor, as above. E-mail: ljt@mcmaster.ca Received 5 December 2013, accepted 1 September 2014 © 2014 Federation of European Neuroscience Societies and John Wiley & Sons Ltd European Journal of Neuroscience, Vol. 40, pp. 3608–3619, 2014 doi:10.1111/ejn.12741