Bryn Mawr Classical Review

BMCR 2013.12.35 on the BMCR blog

Bryn Mawr Classical Review 2013.12.35

John T. Koch, Barry W. Cunliffe (ed.), Celtic from the West 2: Rethinking the Bronze Age and the Arrival of Indo-European in Atlantic Europe. Celtic studies publications, 16.   Oxford; Oakville, CT:  Oxbow Books, 2013.  Pp. viii, 237.  ISBN 9781842175293.  $70.00.  

Reviewed by Joseph F. Eska, Virginia Polytechnic Institute and State University (


This volume is a follow-up to a previous one which discusses ideas of the editors which place the formation of the Celtic language family in the Atlantic zone of Europe in the Bronze Age, as opposed to the traditional view of it having formed in the northern European Alps in the Iron Age.1 Barry Cunliffe, in a number of publications going back to 2001, has sought to show that the spread of Celtic speech throughout Europe from west to east, rather than east to west, squares better with the archeological evidence, while John T. Koch has sought to show that the Celtic language family formed in the Iberian Peninsula and spread northward and eastward, and, more specifically, that the fragmentarily attested Tartessian language attested in the far southwest of the Iberian Peninsula is, in fact, Celtic.2

In his prologue, ‘Ha C1a ≠ PC (“The earliest Hallstatt Iron Age cannot equal proto-Celtic”)’ (p. 1–16), John T. Koch states that, at the beginning of recorded history, all linguistic records are consistent with a Celtic analysis and asserts that Celtic goes back to the Bronze Age, that Tartessian as Celtic is an ‘inescapable’ observation (p. 5), and that that movement of key sword types from west to east supports the notion that Celtic spread from the Iberian Peninsula (p. 6). He is correct when he states that numerous scholars before him have discussed the possibility of Tartessian being Celtic (p. 6), though most of these expressed the idea only tentatively and some have come to reject it. More recently, Javier de Hoz and the present reviewer are firmly of the view that it is not Celtic, while Patrick Sims-Williams labels the opinion that it is Celtic as ‘a minority view’.3 And, as noted in the contributions by Dirk Brandherm and Dagmar S. Wodtko, material change does not imply linguistic change.

J. P. Mallory begins his ‘The Indo-Europeanization of Atlantic Europe’ (p. 17–39) by noting that it is axiomatic that, in the discussion of the spread of languages, linguistic evidence must always take precedence over archeological evidence. He provides a sound critique of how it would be hard to explain the close relationship of the Italic and Celtic languages if the Celtic homeland were in the Iberian Peninsula (p. 24–26) and then marshalls archeological arguments against the idea (p. 27). Mallory also ventures into the area of the lexicon, noting, for example, that the Celtic of Gaul, Britain, and Ireland all share a common etymon for ‘iron’, which would, indeed, be surprising had they separated as early as the Bronze Age. He concludes that Celtic was part of a northwestern group of Indo-European languages and that Celtic arose north of the Pyrenees.

A. P. Fitzpatrick, in discussing ‘The arrival of the Beaker Set in Britain and Ireland’ (p. 41–70) in the second half of the third millennium BCE, adopts the emerging view that it arose in the Iberian Peninsula, as opposed to the lowland countries, as has been the traditional view, and that it was transported to the British Isles directly from the Atlantic zone. He concludes that the mobility of the Bell Beaker culture may have led to the dispersal of languages as has been proposed by Barry Cunliffe.

In ‘Beakers into Bronze; tracing connections between western Iberia and the British Isles 2800–800 BC’ (p. 71–99), Catriona Gibson discusses long-distance interconnections between the Iberian Peninsula and Britain and Ireland via the movement of small groups of people in the Copper and, especially, the Bronze Age, focussing upon objects such as swords, feasting objects, and ornaments. She presents evidence for shared economic, ideological, and symbolic structures and wonders whether, as she puts it, they were ‘imparted through common linguistic frames of reference’ (p. 91).

In ‘Out of the flow and ebb of the European Bronze Age: heroes, Tartessos, and Celtic’ (p. 101–146), John T. Koch continues to advance the idea that Tartessian is a Celtic language and, in doing so, embraces the idea that proto-Celtic did not descend directly from proto-Indo-European, but arose from the coalescence of a variety of Indo-European dialects in the Iberian Peninsula during the Bronze Age. Once formed, it came into contact with non-Indo-European Iberian, which leads Koch to conclude that Celtic is Iberianised Indo-European (p. 137). This does not hold water, however, for, as discussed by Don Ringe (with references) in an informative posting to the Language Log on The linguistic diversity of aboriginal Europe, it is known from the study of bilingual children that ‘[e]xtensive structural convergence of languages, as opposed to mere word-borrowing or the adoption of a few superficial traits, turns out to be rare. … By contrast, dialects that are mutually intelligible can and do merge …’ What this means is that any convergence of dialects that resulted in proto-Celtic would have had to have been so similar as to have descended simply from an earlier phase of proto-Celtic. Moreover, as Ringe notes, ‘divergence or death,’ not convergence, ‘is the normal fate of languages.’ As for Koch’s view that Celtic is Iberianised Indo-European, the only feature that he cites is the loss of the proto-Indo-European */p/ phoneme, which he claims is the result of contact with /p/-less Iberian. But reflexes of proto-Indo-European */p/ do occur in Celtic,4 and the now significant remains of the ancient Continental Celtic languages are noteworthy in that they resemble other ancient Indo-European languages in their structures to a significantly greater extent than the later attested Celtic languages of the Britain and Ireland.

Dirk Brandherm begins ‘Westward ho? Sword bearers and all the rest of it …’ (p. 147–155) by tacitly accepting Koch’s identification of Tartessian as a Celtic language and then seeks to identify evidence in the archeological record to support it. He particularly examines the depiction of weaponry in some of the Tartessian stelae. He notes that there is evidence for movement of such objects from central Europe to the Iberian Peninsula in the Early Bronze Age, and then in the reverse direction in the Late Bronze Age, thinking that such movements ‘might have been accompanied by language shift’ (p. 153). On the basis of just this evidence, he concludes that which directional shift might be associated with the expansion of Celtic through Europe — I note that there is no reason to assume that either was — cannot be based upon current archeological evidence.

Jacqueline I. McKinley, Jörn Schuster, and Andrew Millard’s ‘Dead-Sea connections: a Bronze Age and Iron Age ritual site on the Isle of Thanet’ (p. 157–183) focusses upon the site of Cliffs End Farm in southeastern Britain, which is notable for its large number of burials of individuals, many of whom had migrated from a very significant distance to the site as determined by strontium/oxygen analysis. Also of note is the variety of mortuary rites: ‘communal and individual graves; human and probably animal sacrifice; excarnation with manipulation and redeposition of partially articulated body parts, and curation of individual skeletal elements; exposure with subsequent canid and possibly avian scavenging of remains, and bleaching or charring of some body parts and skeletal elements’ (p. 169). The authors conclude with some remarks on the cultural background tying the grave goods to the Atlantic zone, but also to Scandinavia and western and central Europe.

Dagmar S. Wodtko’s ‘Models of language spread and language development in prehistoric Europe’ (p. 185–206) is an accessible and nuanced account of how the Indo-European languages may have spread. Among her important points are that linguistic and cultural changes/shifts need not correlate with one another. Indeed, when working on such a problem as the spread of the Indo-European languages, it is only linguistic facts which bear upon the resolution of the problem. As Wodtko aptly puts it, ‘the introduction of Indo-European in western Europe could be fairly unspectacular in terms of archaeological visibility’ (p. 200).5

In ‘Early Celtic in the west: the Indo-European context’ (p. 207–217), Colin Renfrew ‘provisionally’ accepts that Tartessian is a Celtic language (p. 209) and proceeds to employ that acceptance to justify his theory that the Indo-European languages dispersed from Anatolia and claims that Indo-European languages were on the Atlantic seaboard by ca. 4000 BCE. By the end of the paper, Koch’s claim that Tartessian is a Celtic language and that Celtic spread throughout Europe from the Iberian Peninsula seems to have become established fact.

Barry Cunliffe’s epilogue, ‘The Celts — where next’ (p. 219–223), states that the view that Tartessian is a Celtic is gaining broad support, a view that, I would contend, is hard to square with the facts. Koch, himself, admits that the language contains ‘[m]any sequences [that] remain profoundly uncertain, lacking any identifiable stem, prefix, or termination; and neither do they form a plausible unit occurring after an identifiable termination and before an identifiable word beginning and/or bounded at one end by the obvious and undamaged beginning or end of the inscribed text’ and that the Tartessian script ‘seems better suited to the phonological structure of the Iberian language’.6 Renfrew claims that all contributors to the volume would accept that Celtic was a lingua franca along the Atlantic coast in the Bronze Age, a statement that I suspect would surprise a number of them.

Jürgen Zeidler, in his BMCR review of the predecessor to this volume, suggests that Tartessian may be a non-Indo-European language containing a large number of Celtic onomastic forms.7 Given that much of the corpus, unlike the undoubtedly Indo-European Celtiberian language, cannot be parsed and that Tartessian phonology appears to be distinctly Iberoid, I am entirely in agreement with such an opinion. Readers should approach the Celtic from the west enterprise, which must rely solely upon linguistic evidence, with the greatest of sceptism.


1.   Barry Cunliffe and John T. Koch (edd.). Celtic from the West: Alternative Perspectives from Archaeology, Genetics, Language and Literature. Oxford: Oxbow Books, 2010. See the BCMR review by Jürgen Zeidler at BMCR 2011.09.57.
2.   For example, Barry Cunliffe, Facing the Ocean: The Atlantic and its Peoples 8000 BC–AD 1500. Oxford: Oxford University Press, 2001, p. 293–297; ‘Celticization from the west: the contribution of archaeology’, in Celtic from the West, p. 13–38; John T. Koch, Tartessian: Celtic in the south-west at the dawn of History. Aberystwyth: Celtic Studies Publications, 2009; Tartessian 2. Aberystwyth: University of Wales Centre for Advanced Welsh and Celtic Studies, 2011.
3.   Javier de Hoz. Historia linguistica de la Península Ibérica en la antigüedad ii, El mundo ibérico prerromano y la indoeuropeización. Madrid: Consejo Superior de Investigaciones Científicas, 2010, p. 588; Joseph F. Eska, reviews of John T. Koch, Tartessian and Tartessian 2, forthcoming in Kratylos 58, 2013; Patrick Sims-Williams, ‘Bronze- and Iron-Age Celtic speakers. What don’t we know, what can’t we know, what could we know? Language, genetics and archaeology in the twenty-first century’. The Antiquaries Journal 92: 1–23.
4.   See recently Joseph F. Eska, ‘In defense of Celtic /φ/’, in Adam I. Cooper, Jeremy Rau, and Michael Weiss (edd.), Multi nominis grammaticus: Studies in Classical and Indo-European linguistics in honor of Alan J. Nussbaum on the occasion of his sixty-fifth birthday. Ann Arbor: Beachstave Press, 2013, p. 32–43.
5.   Cf. David W. Anthony, The Horse, the Wheel, and Language: How Bronze-Age Riders from the Eurasian Steppes Shaped the Modern World. Princeton: Princeton University Press, 2007, p. 117–119. He discusses how small elite groups can motivate language change and recounts an ethnohistorical study of the Acholi, an ethnolinguistic group in northern Uganda and southern Sudan, whose language, Luo, became dominant even though the initial number of immigrants into the area were few.
6.   Koch, Tartessian 2, p. 87–88 and 164–165, respectively.
7.   NB that Jürgen Untermann, in his primary edition of the Tartessian corpus, Monumenta linguarum Hispanicarum iv, Die tartessischen, keltiberischen und lusitanischen Inschriften. Wiesbaden. Dr. Ludwig Reichert, 1997, notes that there appears to be a strict delineation in the lexicon of Tartessian between onomastic forms, which often appear to be Celtic, or at least Indo-European, and the non-onomastic part, which looks distinctly different (p. 167).

Read comments on this review or add a comment on the BMCR blog

Read Latest
About BMCR
Review for BMCR
Support BMCR

BMCR, Bryn Mawr College, 101 N. Merion Ave., Bryn Mawr, PA 19010