LINGUIST List 14.2178

Mon Aug 18 2003

Review: Historical Linguistics: Andersen, ed. (2003)

Editor for this issue: Tomoko Okuno <>

What follows is a review or discussion note contributed to our Book Discussion Forum. We expect discussions to be informal and interactive; and the author of the book discussed is cordially invited to join in. If you are interested in leading a book discussion, look for books announced on LINGUIST as "available for review." Then contact Simin Karimi at


  1. Marc L Greenberg, Language Contacts in Prehistory: Studies in Stratigraphy

Message 1: Language Contacts in Prehistory: Studies in Stratigraphy

Date: Sat, 16 Aug 2003 14:57:49 +0000
From: Marc L Greenberg <>
Subject: Language Contacts in Prehistory: Studies in Stratigraphy


Andersen, Henning, ed. (2003) Language Contacts in Prehistory: Studies
in Stratigraphy (Papers from the Workshop on Linguistic Stratigraphy
and Prehistory at the Fifteenth International Conference on Historical
Linguistics, Melbourne, 17 August 2001), John Benjamins Publishing
Company, Current Issues in Linguistic Theory 239.

Announced at

Marc L. Greenberg, University of Kansas


The book under review presents papers from a workshop on linguistic
stratigraphy as part of the Fifteenth International Conference on
Historical Linguistics in Melbourne in 2001. The term ''stratigraphy''
and related terms (substratum, adstratum, etc.) are metaphors
originally corresponding to geological referents, also used in
archaeology. Here they refer to the layering of linguistic material
whether developed through internal innovation or acquired in contact
situations in the prehistoric past. The case is made that this
linguistic evidence is often more eloquent than archaeological data
for the same time frame, even when the two can be correlated with one
another. Specific problems and advances in understanding both
individual language situations and broader theoretical issues are
discussed in conjunction with the examination of material from Indo-
European, African, Southeast Asian, Australian, Oceanic, Japanese, and
Meso- American languages.


A brief Preface (v - vi) mentions the circumstances under which
Henning Andersen and Christopher Ehret developed the idea to hold a
workshop on the theme of linguistic stratigraphy at the 15th ICHL.

The chapter Introduction (1 - 10) by Henning Andersen begins with a
discussion of the metaphors surrounding ''stratigraphy'' (from
geology, and, priorly, metallurgy and chemistry) as employed in
language reconstruction -- stratum, substratum, and superstratum. The
tentative nature of the application of stratigraphic metaphors in
linguistic reconstruction is suggested by Andersen's comments on the
lack of standard diagrams: ''Some historical linguists who work with
dialect data make more schematic time-- space diagrams informally for
teaching purpose. But they do not often occur in print [...] and are
not described in standard handbooks of dialectology, which generally
limit themselves to surface maps showing signatures and/or isolines
[...].'' The second part of the chapter is devoted to Language
Contact in Prehistory, which discusses in some depth the notions
''borrowing'' and ''intrusion'' (like borrowing but not pragmatically
motivated), as well as ''transfer'' and ''interference'' as two types
of intrusion. In a section on Typologies of Contact Changes Andersen
takes issue with the formulation of Thomason's 1997 typology of
''mechanisms'' and ''processes'' of ''interference'' features. Rather
than seeing ''code-switching'' and ''code-alternation'' as
''mechanisms,'' Andersen describes them as bilingual and diglossic
settings in which borrowing, transfer, and interference may
occur. Andersen proposes ''metadialogue'' corresponding to Thomason's
''negotiation,'' which he describes using the metaphor of legislation:
speakers in linguistic community entertain a motion and decide whether
the motion (accession, innovation) is carried (actualized) or rejected
in the community's norms. Crucially for stratigraphy, Andersen points
out that whether or not in the aftermath it is possible to determine
the nature of an accession -- borrowing, transfer, interference -- the
distinctions among the types still hold.


In Stratum and Shadow: A Genealogy of Stratigraphy Theories from the
Indo-European West (11 - 44), Bernard Mees treats the particular
burdens of European linguistic prehistory, in which nineteenth-century
(or earlier) preconceived notions of race and the knowledge of names
of earlier populations are found to have driven some theories of
substrata. Ironically, (invariably imperfect) historical documentation
is seen by non-Indo- Europeanists as a tacit advantage over doing
prehistorical reconstruction on non-literate traditions (cf. McConvell
and Smith on the prehistory of the indigenous languages of Australia,
p. 183). Mees identifies the break with nineteenth-century
historicism, which saw in the genealogical approach a past to a
present system, in the work of the W�rter-und- Sachen
theorists, who view the linguistic past as past systems. Prior to the
W�rter-und- Sachen theorists, however, stratigraphic approaches
were seen in ethnic and cultural, rather than linguistic, terms and,
according to Mees, much of the thinking about substratal issues in
Western Europe goes back to the work of the French Celticist Henri de
Jubainville (1827 - 1910), the originator of the notion of the Gaulish
substratum to French. Many of these theories, he asserts, have
''little claim to a proper empirical basis today as they were first
proposed in light of analyses of the now-outdated philological record
available in the late nineteenth century'' (13). For example, the
fronting of Latin /u/ to /y/ and lenition of voiceless
stops are traditionally attributed to the Gaulish substratum. Yet,
fronting of /u/ is unattested in Continental Celtic and Gaulish Latin;
moreover, lenition in Gaulish is known to have been partial and what
earlier scholars had adduced as evidence of lenition is ambiguous,
attributable in part to the imprecision of Latin writing
practice. Mees convincingly shows these and other changes to be
ascribable more complex processes, in some cases to Sprachbund
effects. Regarding shadows, Mees treats some of the more tenuous
attempts -- often confidently asserted, repeated and continued even in
some present-day work -- at identifying linkages between names of
ancient peoples (Ligurians, Venetes, Illyrians) and the extremely
parsimonious evidence available about ''their'' erstwhile
languages. Moreover, these ethnically motivated theories had some of
their darkest days on both sides of Nazi- era racial theory. (A good
review of corresponding Eastern-European linguistic- and
archaeology-based ethnogenetic theories may be found in Curta 2002.)

Henning Andersen's Slavic and the Indo- European Migrations (45 - 76)
first presents a disciplined though brief account of the multifarious
lexical accessions in Baltic and Slavic. The author proceeds from the
model of the relationship between Baltic and Slavic elaborated in
Andersen 1996, in which, though these Indo-European dialects may have
diverged at first, they later constituted a geographical continuum
subsequently obscured by the partial shift towards Slavic at the
expense of Baltic. The meat of the paper is in Andersen's attempt to
sort out five enduring and recalcitrant issues of Baltic and Slavic
historical phonology, each connected with an as yet inadequately
explained set of discrepancies in the reflexes: (i) Baltic st for
Proto-Indo- European (PIE) *k'; (ii) Slavic and Baltic velar stops for
PIE *k', *g'(h); (iii) Slavic and Baltic uR diphthongs for PIE
syllable *r; (iv) Slavic and Baltic e- for PIE *h2e- and *h3e-; (v)
Slavic k- for PIE *h2-, *h3-. Andersen wrings out of the fragmentary
evidence more than had been previously by relating the issues to one
another, as well as to other, better understood changes affecting
related phonological domains, a technique characteristic of several of
his earlier works. He concludes that the data suggest successive waves
of Indo-European settlers: (ii - v) may be attributable to centum
contact, (v) is possibly due to contact with pre-IE; (ii) and (iii)
may include centum and pre-satem dialect contact, a distinction that
may be irrelevant to (iv, v). The relative chronology of accessions
are from the most recent (i) to the earliest (v).

Bridget Drinka's paper, The Development of the Perfect in Indo-
European: Stratigraphic Evidence of Prehistoric Areal Influence (77 -
105), deals not with language contact but with issues of
stratification of linguistic innovation in Indo-European, including
contact between related dialects. The author illustrates this issue on
basis of the development of the perfect, its stratigraphic layering
and areal diffusion as in index to the geographical place of dialects
emerging from the proto-language. Through a close examination of Greek
and Sanskrit evidence, as well as a look at developments elsewhere,
she makes the case for a dynamic innovative area of contact between
pre- Indo-Iranian and pre-Greek in the east in the last stage of the
development of PIE. The overarching implication of the paper is that
PIE was not a uniform proto-language, but one that had already
developed dialect differentiation before its geographic dissolution.


Two articles on African problems are introduced by Christopher Ehret's
comments in Stratigraphy in African Historical Linguistics (107 -
114), including the same author's article on Nilo-Saharan (see
below). These two articles present topics of the greatest complexity
in the volume both with respect to the multifaceted and numerous
language contacts as well as the lengthy time depths concerned. Ehret
points out that these stratigraphic studies indicate the more advanced
state of the art in east- as opposed to west-African historical

Stratigraphy and Prehistory: Bantu Zone F (115 - 134) by B.F.Y.P.
Masele and Derek Nurse deals with the enormous complexity of
establishing genetic relationships among a subgroup of Bantu languages
that had migrated into an area of east-central Africa and undergone
millennia of contacts with non- Bantu languages and each other. The
complexity is compounded by the lack of reliable data on many of the
languages and dialects, a shortcoming that has only recently begun to
be ameliorated. Using lexicostatistics and detailed analysis of key
phonological processes, the authors establish a core genetic grouping
for Bantu Zone F, consisting of five or, possibly, six of the
languages and exclude some others. The analysis supports, by and
large, the scenario sketched in an earlier work by Ehret 1998.

In Language Contacts in Nilo-Saharan Prehistory (135 - 157)
Christopher Ehret examines the contacts of the Rub languages, a subset
of Nilo- Saharan languages once spoken in southern Sudan, northeastern
Uganda, and parts of Kenya and Tanzania, over a period of five to six
millennia. Parallel to the problems inherent in the Bantu problem of
the previous chapter, part of the difficulty here is in teasing out
the various strains of Nilo- Saharan in contact with the Rub
subgroup. The author summarizes the results of a complex, but
relatively clear (as compared to those in the Bantu problem) set of
interrelated sound changes in the relevant language groups in a
Stammbaum-like diagram of the relationships (a diagram mislabeled as
''Figure 1'' but referred to in the text as ''Figure 5'' [pp. 150 -

Southeast Asia

Anthony Diller's paper on Evidence for Austroasiatic Strata in Thai
(159 - 175) investigates the interaction between Mon Khmer and other
Austroasiatic languages with Thai in an attempt to stratify these
contacts. Because of the protracted periods of contact it is difficult
to distinguish between direct contact and Sprachbund phenomena. For
this reason, the paper attempts to establish an eclectic and
probabilistic methodological approach to Thai stratigraphy. The author
relies upon available inscriptional and lexicographical resources to
develop distributional hypotheses for further testing. As in several
other papers in the collection, the testing involves correlation with
extra-linguistic evidence.


In Patrick McConvell and Michael Smith's paper, Millers and Mullers:
The Archaeo- Linguistic Stratigraphy of Technological Change in
Holocene Australia (177 - 200), the authors argue for a tight
interdisciplinary cooperation between archaeology and linguistics in
tackling language reconstruction of preliterate society at several
thousand years before the present. In doing so they highlight recent
work that they have done on both the archaeology (Smith) and
linguistics (McConvell) of seed-grinding technology as this technology
improved and diffused throughout the native population of Australia.


Hans Schmidt's Loanword Strata in Rotuman (201 - 240) studies the
stratification of loans into Rotuman, a Central Pacific language of
Oceanic, which, owing to its complex morphophonology and peculiar
sociolinguistic situation poses special challenges to historical
reconstruction. Through close analysis of phonology, morphology, and
historical (including mythological) information, the author builds a
stratigraphy of loans into Rotuman from the thirteenth century to the
present, associating the various source languages with the semantic
fields that they contributed and the era relevant to the contact
(e.g., Eastern U'vean corresponding to social stratification
vocabulary, chiefly language and titles, dating to the Tongan empire,
from the sixteenth to the beginning of the nineteenth century).


J. Marshall Unger's paper, Substratum and Adstratum in Prehistoric
Japanese (241 - 258), discusses the position of Japanese with respect
to Korean, which has implications also for a potential widening of the
Altaic family or an even larger grouping. Unger argues for a careful
and patient bottom-up approach to grouping (characteristic of his own
and Samuel E. Martin's view of the relationship of Korean and
Japanese) and against top-down theories (such as those of Roy Andrew
Miller and Joseph H. Greenberg). Unger points out that the main
problem with positing a Japanese-Korean common starting point is that
the two languages have fewer cognates than would be expected for a
relatively short 2,300-year separation. The author's solution is an
adstratum hypothesis. He proposes that Proto-Korean-Japanese (his
proto-Samhan- Wa) entered the southern Korean peninsula and Northern
Kyushu around the third century B.C., perhaps from the Chinese coast
between the mouth of the Yangzi and the Shandong peninsula. This
population linguistically assimilated the previous populations of the
two areas as they spread southward in the peninsula and the island,
respectively. In the Korean peninsula the languages of the Koguryo and
Puyo (who established the kingdom of Paekche), which have affinity
with Tungusic, were influential both in Korea and, as newcomers, in
Japan. Prestige lexical items from the language of the newcomers
displaced much native vocabulary in Japan, but not in Korea, parallel
to the partial displacement by French of Anglo-Saxon lexis. For this
reason, many proto-Korean-Japanese words that were preserved in Korean
are displaced in Japanese. The stratigraphic evidence for this
scenario is found in the matching of etyma in the following way: a
special set of Korean- Japanese etymologies involve an Old Japanese
word with a limited distribution or semantic range lacking a cognate
in Korean. These etyma match words known from vestigial evidence of
the languages of Koguryo or Paekche on the one hand, or Tungusic
languages on the other.


In Uto-Aztecan in the Linguistic Stratigraphy of Mesoamerican
Prehistory (259 - 288), Karen Dakin uses linguistic stratigraphic
methods to tease out evidence for whether or not Uto- Aztecan,
especially Nahuatl, was relatively early or late arrival into Meso-
America. In particular, the author examines the special features of
Nahuatl word-formation, especially compounding, as the words were
borrowed or calqued into Meso-American languages. It is found that the
presence of early or even pre-Nahuatl forms, preceding known sound
changes within the language, points to earlier contact than had been
suspected by archaeologists and ethnohistorians.

A Language Index (289 - 292) is provided at the end.


The volume presents excellent contributions examining several major
world areas, affording a glimpse into the problems, tasks, and
relative successes in applying stratigraphic techniques to language
contact situations in prehistory. The contributions demonstrate that
linguistic stratigraphic techniques can be successfully integrated
with archaeological and other extra- linguistic to maximize the
robustness of explanation, though in virtually all the papers the
linguistic approach is shown to be superior. It strikes this reviewer
that the relative dearth of historical textual evidence in
non-Indo-European families ratchets up the level of ingenuity brought
to bear on the problem of prehistoric contact. Researchers of
linguistic prehistory will find this volume very useful to glean not
only the commonalities and best-practices used to extract the most
explanation out of complex and often recalcitrant prehistorical
linguistic data, but also learn a few new tricks by looking over the
fence at what researchers in specialized areas have developed to deal
with the particular (arguably, unique) complexities in their own


Andersen, Henning. 1996. Reconstructing Prehistorical
Dialects. Initial Vowels in Slavic and Baltic (= Trends in
Linguistics, Studies and Monographs 91). Berlin/New York: Mouton de

Curta, Florin, 2002. From Kossina to Bromley: Ethnogenesis in Slavic
Archaeology. Andrew Gillett, ed. On Barbarian Identity. Critical
Approaches to Ethnicity in the Early Middle Ages (= Studies in the
Early Middle Ages, v. 4): 201 - 218. Turnhout, Belgium: Brepols.

Ehret, Christopher. 1998. An African Classical Age. Charlottesville:
University Press of Virginia.

Thomason, Sarah G. 1997. On Mechanisms of Interference. Language and
its Ecology. Essays in Memory of Einar Haugen. Stig Eliasson & Ernst
Hakon Jahr, eds: 181 - 208. Berlin/New York: Mouton de Gruyter.

Marc L. Greenberg (UCLA Ph.D., 1990), Professor and Chair of the
Department of Slavic Languages and Literatures at the University of
Kansas, specializes in Slavic historical linguistics and dialectology.
He is a co-founding editor of Slovenski jezik / Slovene Linguistic
Studies. His book, A Historical Phonology of the Slovene Language
(2000) in the Carl Winter series, was recently translated into Slovene
and published by Aristej Publishing Co., Maribor.
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue