I'm doing a little literature review this week on Middle Pleisocene postcrania. On a somewhat tangential topic, the description of the Sima de los Huesos cervical vertebrae, by Gómez-Olivencia and colleagues (2007), includes a nice summary of the current knowledge of the thoracic vertebral canal of KNM-WT 15000 and other early Homo specimens.
Much attention has been devoted to vertebral-canal size and its relationship to spoken language. One factor in the evolution of human language that would be reflected in vertebral-canal morphology is increased breath control (MacLarnon, 1993, MacLarnon and Hewitt, 2004). Modern humans have an enlarged thoracic vertebral canal, reflecting a larger amount of gray matter. Based on the morphology of the KNM-WT 15000 individual, a narrower thoracic canal has been proposed for Homo ergaster, indicating that this species may only have been capable of short, unmodulated utterances, such as those used by extant nonhuman primates (MacLarnon and Hewitt, 1999). However, significant abnormalities have been found in the KNM-WT 15000 individual (Latimer and Ohman, 2001), which could indicate some form of axial dysplasia, and so the small canal may be a reflection of a neural-canal stenosis associated with the pathology. In contrast, Schiess et al. (2006) argued that the diagnosis of a congenital dysplasia is not supported, indicating that the pathological lesions in the KNM-WT 15000 individual may not be as severe as previously reported. Moreover, the Dmanisi vertebrae (Meyer, 2005 and Meyer et al., 2006), which are the oldest known for the genus Homo, follow the modern human pattern in all regions, as the raw and relative sizes of the vertebral canals fall well within the human range, indicating that these hominins may have had fine control of the respiratory muscles involved in spoken language (Meyer, 2005 and Meyer et al., 2006).
Arsuaga et al. (1997a) showed that the mean cranial capacity of SH's three most complete crania (1245 cm3) (Arsuaga et al., 1993 and Arsuaga et al., 1997c) is slightly less than that of two comparative samples from the Hamann-Todd Osteological Collection. However, given the large body-weight estimates for these hominins, their encephalization quotients are below both modern human or Neandertal values (Arsuaga et al., 1999). In Neandertals, higher encephalization quotients are reached by expansion of the cranial capacity, while in modern humans it is mainly achieved by a reduction in body mass (Arsuaga et al., 1999 and Carretero et al., 2004). In addition to the parallel trends in encephalization in these two lineages, the absolute size of the bony vertebral canal in the upper cervical spine reached modern human values by the middle Pleistocene. Preliminary studies (Carretero et al., 1999, Gmez et al., 2004 and Gmez-Olivencia, 2005) have shown that the SH lower cervical spine's canal had a similar size compared to modern humans, but a full assessment of this anatomical region will not be possible until larger sets of cervical and thoracic vertebrae are associated. In any case, as demonstrated by Martnez et al. (2004), the SH hominins had the skeletal characteristics of the outer and middle ear that support the perception of spoken language (Gómez-Olivencia et al. 2007:22).
The Meyer references are to Marc Meyer's dissertation on the Dmanisi vertebral remains and a subsequent conference presentation. I think those are more than sufficient to say that this particular piece of anatomy isn't evidence for restricted breathing control in early Homo. I don't have much more to say, just though these two paragraphs sum up a lot of information in a useful way.
Gómez-Olivencia A, Carretero JM, Arsuaga JL, Rodríguez-García L, García-González R, Martínez I. 2007. Metric and morphological study of the upper cervical spine from the Sima de los Huesos site (Sierra de Atapuerca, Burgos, Spain). J Hum Evol 53:6-25. doi:10.1016/j.jhevol.2006.12.006