Speech capabilities in Homo erectus: Dmanisi & Nariokotome vertebrae compared

[Daryl Habel]

Catch this from Science News Online,
Week of May 6, 2006; Vol. 169, No. 18 , p. 275
 

Evolutionary Back Story: Thoroughly modern spine supported human ancestor
Bruce Bower

Bones from a spinal column discovered at a nearly 1.8-million-year-old site in central Asia support the controversial possibility that ancient human ancestors spoke to one another. Excavations in 2005 at Dmanisi, Georgia, yielded five vertebrae from a Homo erectus individual, says anthropologist Marc R. Meyer of the University of Pennsylvania in Philadelphia. The finds occurred in previously dated sediment that has yielded several skulls now attributed to H. erectus (SN: 5/13/00, p. 308: Available to subscribers at http://www.sciencenews.org/articles/20000513/fob1.asp).

The new discoveries represent the oldest known vertebrae for the genus Homo, Meyer announced last week at the annual meeting of the Paleoanthropology Society in San Juan, Puerto Rico. The fossils consist of one lumbar, two thoracic, and two cervical vertebrae.

Meyer and his colleagues—David Lordkipanidze and Abesalom Vekua, both of the Georgian State Museum in Tbilisi—compared the size, shape, and volume of the Dmanisi vertebrae with more than 2,200 corresponding bones from people, chimpanzees, and gorillas.

"The Dmanisi spinal column falls within the human range and would have comfortably accommodated a modern human spinal cord," Meyer says.

Moreover, the fossil vertebrae would have provided ample structural support for the respiratory muscles needed to articulate words, he asserts. Although it's impossible to confirm that our prehistoric ancestors talked, Meyer notes, H. erectus at Dmanisi faced no respiratory limitations on speech.

In contrast, the 1984 discovery in Kenya of a boy's 1.6-million-year-old skeleton, identified by some researchers as H. erectus and by others as Homo ergaster, yielded small, chimplike vertebrae. Researchers initially suspected that the ancient youth and his presumably small-spined comrades lacked the respiratory control to talk as people do today.
Excavations in 2005 at Dmanisi, Georgia, yielded five vertebrae from a Homo erectus individual, says anthropologist Marc R. Meyer of the University of Pennsylvania in Philadelphia. The finds occurred in previously dated sediment that has yielded several skulls now attributed to H. erectus (SN: 5/13/00, p. 308: Available to subscribers at http://www.sciencenews.org/articles/20000513/fob1.asp).

The new discoveries represent the oldest known vertebrae for the genus Homo, Meyer announced last week at the annual meeting of the Paleoanthropology Society in San Juan, Puerto Rico. The fossils consist of one lumbar, two thoracic, and two cervical vertebrae.

Meyer and his colleagues—David Lordkipanidze and Abesalom Vekua, both of the Georgian State Museum in Tbilisi—compared the size, shape, and volume of the Dmanisi vertebrae with more than 2,200 corresponding bones from people, chimpanzees, and gorillas.

"The Dmanisi spinal column falls within the human range and would have comfortably accommodated a modern human spinal cord," Meyer says.

Moreover, the fossil vertebrae would have provided ample structural support for the respiratory muscles needed to articulate words, he asserts. Although it's impossible to confirm that our prehistoric ancestors talked, Meyer notes, H. erectus at Dmanisi faced no respiratory limitations on speech.

In contrast, the 1984 discovery in Kenya of a boy's 1.6-million-year-old skeleton, identified by some researchers as H. erectus and by others as Homo ergaster, yielded small, chimplike vertebrae. Researchers initially suspected that the ancient youth and his presumably small-spined comrades lacked the respiratory control to talk as people do today....(more)...

That's the first half of Bruce Bower's Science News Online story.  The contention that Homo erectus lacked modern speech capability because the neural canal in the spinal vertebrae of WT 15000 (the Nariokotome Boy) was considerably smaller than in modern humans was made by Alan Walker and his colleagues during the latter part of the last century.  The neural canal in the recently discovered Dmanisi vertebrae, however, are of modern human dimensions.   Meyer, Lordkipanidze & Vekua say this supports the idea that the Nariokotome Boy suffered from "a developmental pathology known as neural stenosis," and therefore there is no reason why Homo erectus was not perfectly capable of normal human speech.
However, in the second half of the Science News Online story by Bruce Bower, there is a disagreement:

Robert C. McCarthy of Florida Atlantic University in Boca Raton disagrees. At the Paleoanthropology Society meeting, he presented vocal-tract reconstructions for various ancient Homo species suggesting that the capacity to articulate speech as well as people do now emerged exclusively in Homo sapiens around 50,000 years ago.

Before then, all members of the Homo genus—including H. sapiens—possessed a short set of neck vertebrae, resulting in a vocal tract with a restricted range of speech sounds, McCarthy and his coworkers argue.

Many populations today, including Australian aborigines, possess neck vertebrae comparable in length to those that McCarthy's team considered inadequate for modern speech, Meyer responds.

The entire Science News Online article, with a nice photographic comparison of the Dmanisi and Nariokotome vetebrae neural canals, can be read at:CLICK HERE

The references given in the above Science News Online story by Bruce Bower derive from the Paleoanthropology Society meeting held 24-26 April 2006 in San Juan, Puerto Rico, and are cut-and-pasted from the abstracts below (note that in the first reference, McCarthy et al., one of the co-authors is Philip Lieberman who, throughout his career, has published extensively on the subject of archaic human speech, generally opposed to the idea of modern human speech capabilities):

McCarthy, R.C., et al. 2006. The origin of human speech. Paleoanthropology Society meeting. April 24. Puerto Rico.

A23
24 April 2006
The Origin of Human Speech
Robert C. McCarthy, Department of Anthropology, Florida Atlantic University
David S. Strait, Department of Anthropology, University at Albany
Franklin Yates, Department of Anthropology, Harvard University
Philip Lieberman, Department of Cognitive and Linguistic Sciences, Brown University
-----
Modern humans have a unique pharyngeal configuration in which the larynx lies well below a short, retracted oral cavity, creating a supralaryngeal vocal tract (SVT) with horizontal and vertical components of approximately equal length. This “two tube” SVT dynamically filters sounds produced by the passage of air through the vocal folds in much the same way that a wind instrument filters air after it has been blown through a reed. When the lengths of the horizontal and vertical components of the SVT are in a one-to-one (1:1) ratio, vowels have highly distinctive and stable spectral peaks that place lesser demands on articulatory precison because their formant patterns remain stable over a range of tongue positions. By melding the formant pattern of individual phonemes into syllables transmitted at a slower rate, such “quantal” vowels reduce perceptual errors by allowing a data transmission rate exceeding the frequency of the human auditory system.
Quantal speech is habitually produced only by humans, and thus it is of great interest when a 1:1 SVT evolved. However, comparative anatomical studies yield conflicting SVT predictions for extinct humans including Neanderthals. Here we take a different approach, assessing the viability of differently-proportioned reconstructions relative to limits imposed on SVT morphology by swallowing and breathing. Results indicate that a 1:1 SVT appears first in the skulls of Upper Paleolithic Homo sapiens, well after the species originated 150-200,000 years ago. Although Middle Paleolithic humans (including both Neanderthals and Skhul V) possessed subequal SVTs, reconstructed spectral and formant characteristics of the vowel (i) (“eee”) suggest that they were capable of producing quantal speech, albeit with less room for error than Upper Paleolithic and modern H. sapiens. This suggests that fully quantal speech capabilities arose relatively recently, and may have played a role in the Upper Paleolithic revolution.

Meyer, M.R., D. Lordkipanidze, and A. Vekua. 2006. Language and empathy in Homo erectus: Behaviors suggested by a modern spinal cord from Dmanisi, but not Nariokotome. Paleoanthropology Society meeting. April 24. Puerto Rico.

A20
24 April 2006
Language and Empathy in Homo erectus: Behaviors Suggested by a Modern Spinal Cord From Dmanisi, But Not Nariokotome
Marc R. Meyer, University of Pennsylvania School of Medicine, Department of Cell and Developmental Biology
David Lordkipanidze, Georgian State Museum, Tbilisi
Abesalom Vekua, Georgian State Museum, Tbilisi
-----
Recently, a partial spinal column attributed to Homo erectus was discovered at the site of Dmanisi, Georgia (Meyer 2005). Dated to 1.78 million years before present, the Dmanisi vertebrae are the oldest known for the genus, and present an important opportunity to examine the spinal anatomy and neuroanatomical potential of early Homo.
Comparative analyses against 2,257 human, chimpanzee and gorilla vertebrae demonstrate that the shape and size (absolute and size-corrected) of the Dmanisi spinal cord in the cervical, thoracic, and lumbar regions would have matched that of modern humans. This contrasts with the only other spinal column known for early Homo, the Nariokotome specimen (KNM-WT-15000), which exhibits evidence for a small and apelike spinal cord. The Nariokotome specimen suggested to some workers that early Homo lacked a human level of innervation to respiratory muscles involved in spoken language, and was therefore, incapable of producing spoken language. However, this study unequivocally supports suggestions that the KNM-WT-15000 vertebrae exhibit a developmental pathology known as neural stenosis, and is not representative of the taxon (Latimer & Ohman 2001; Meyer 2003). Moreover, this pathological condition may have conferred chronic health and locomotor difficulties in the Nariokotome individual, precipitating a considerable degree of assistance from conspecifics. In contrast, a quarter-million years before the birth of Nariokotome, Homo erectus at Dmanisi had already evolved a fully human postcranial neuroanatomical substrate associated with the control of respiratory muscles involved in spoken language. Thus, both altruism and spoken language may have been part of the behavioral repertoire of early Homo.
------
Latimer, B, and Ohman, J. 2001. Axial Displasia in Homo erectus. Journal of Human Evolution 40: A12.
Meyer, M. 2003. Vertebrae and Language Ability in Early Hominids. PaleoAnthropology 1: 20–21.
Meyer, M. 2005. Functional Anatomy of the Homo erectus Axial Skeleton from Dmanisi, Georgia. Ph.D. thesis. Philadelphia: Department of Anthropology, University of Pennsylvania.

Dar

[AWSX]

From what I have read elsewhere there is also some evidence of a social structure at Dmanisi. One article speculated that the toothless old man would not have survived without some social network to care for him. That implies some form of communication however primative.
http://pharyngula.org/index/weblog/comments/old_man_of_georgia/

[shenzhou]

Interesting stuff, i seem to recall that some pathology in the spine of  WT 15000 was recognised quite a long time ago although i can't remember exactly what it was.

WT 15000 also apparently has a brocas cap visible in his brain endocast and it has human like asymmetry, this suggests that it had some sort of ability to speak.

Here is the abstract of Marc R Meyer's dissertation on this subject:

"Functional biology of the Homo erectus axial skeleton from Dmanisi, Georgia

 Abstract

A new partial skeleton of Homo erectus attributed to the D2700 cranium was recently discovered at the site of Dmanisi, Georgia. Two cervical, two thoracic, and one lumbar vertebra are among these remains, and are also attributed to this individual. The vertebrae represent the oldest known vertebral series for this taxon, and are the oldest known outside of Africa. Because elements from each spinal region are represented, including never before seen cervical elements for this species, these fossils provide an important look at the axial skeleton of Homo erectus.

The fossils were compared 2257 individual vertebrae from modern humans, chimpanzees, gorillas and fossil hominids to assess the morphological and functional structure of the Dmanisi vertebrae and associated spinal cord.

Although the vertebrae correlate with a stature of only 1.39 meters (4&feet; 7&inches;), the Dmanisi vertebrae reveal key changes relative to earlier small-bodied hominids, with an anatomy for efficient long-range travel and an increased ability to bear compressive loads. Hominid range expansion beyond the African continent is, therefore, associated with not only a shift in subsistence pattern, but also with development of the ability to travel long distances with children (both in utero and in infancy).

Both raw and relative values reveal that the Dmanisi spinal cord in all vertebral regions would have been fully modern in its shape and size. This contrasts with previous studies suggesting that brain evolution outpaced spinal cord evolution in the hominid lineage. The anomalous Nariokotome Homo erectus vertebrae appear to reflect pathology related to infant malnutrition, explaining why its vertebral canals and associated spinal cord size small, approximating australopithecine and chimpanzee values.

These results imply that early Homo erectus departed from the ancestral continent with a postcranial neurological substrate affording high potential for throwing projectiles accurately, and for highly coordinated manual tasks, relative to their australopithecine predecessors. Moreover, this study shows that Homo erectus would have had no postcranial neuromuscular restrictions on its ability to control respiration, and, as such, possessed a fully human postcranial anatomy associated with control and coordination of respiratory muscles for spoken language."

http://repository.upenn.edu/dissertations/AAI3165730/ - sadly at $ 35 for the pdf it's a bit expensive.

[trehinp]

This could also chalenge the theory about the origins of language that Stephen Mithen develops in one of his very recent books : “The Singing Neanderthal: The  Origins of Music, Language, Mind and Body”, Weidenfeld and Nicolson, London 2005

I just finihed reading that book and I have more questions about Mithen's theory than would be reasonable to expose inside this discussion forum. In my analysis, these atricles about Dmanisi Homo Erectus contradict at least some of the argumentation in “The Singing Neanderthal".

By the way if some of the Palanth forum would be interested to start a discussion about that book I would be very pleased. We could start another thread.

Paul Trehin

[lagarvelho]

Paul:

I would be happy to have you or anybody else start a thread about this book.  I will have to start reading it, though!  But if somebody wants to start such a thread, I'll be happy to contribute or ask questions once I've started reading.
Anne G